final_visualise_glacier_projections_until_2300_lowess_percentile_intervals.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "e86b36be",
   "metadata": {},
   "source": [
    "# Notebook to visualise the glacier projections until 2300 of different large-scale glacier models \n",
    "## The created figures are for the ICCI report 2023"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "9a9fd1c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# We use the moepy package to compute lowess prediction intervals with bootstrapping methods\n",
    "# but we don't need it for this notebook, as we precomputed them with \n",
    "# the notebook \"glacier-model-standard-projections/lowess_percentile_interval_fit_per_region.py\"\n",
    "# to use moepy, do: \n",
    "# pip install moepy\n",
    "# https://ayrtonb.github.io/Merit-Order-Effect/\n",
    "# from moepy import lowess, eda"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "7f90928e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import xarray as xr\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import seaborn as sns\n",
    "import os\n",
    "matplotlib.rcParams['figure.figsize'] = (14, 8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "ec370959",
   "metadata": {},
   "outputs": [],
   "source": [
    "# region descriptions ...\n",
    "d_reg_num_name = {}\n",
    "d_reg_num_name['global'] = 'Global'\n",
    "d_reg_num_name['01'] = 'Alaska'\n",
    "d_reg_num_name['02'] = 'Western Canada & USA'\n",
    "d_reg_num_name['03'] = 'Arctic Canada North'\n",
    "d_reg_num_name['04'] = 'Arctic Canada South'\n",
    "d_reg_num_name['05'] = 'Greenland'\n",
    "d_reg_num_name['06'] = 'Iceland'\n",
    "d_reg_num_name['07'] = 'Svalbard and Jan Mayen'\n",
    "d_reg_num_name['08'] = 'Scandinavia'\n",
    "d_reg_num_name['09'] = 'Russian Arctic'\n",
    "d_reg_num_name['10'] = 'North Asia'\n",
    "d_reg_num_name['11'] = 'Central Europe'\n",
    "d_reg_num_name['12'] = 'Caucasus and Middle East'\n",
    "d_reg_num_name['13'] = 'Central Asia'\n",
    "d_reg_num_name['14'] = 'South Asia West'\n",
    "d_reg_num_name['15'] = 'South Asia East'\n",
    "d_reg_num_name['16'] = 'Low Latitudes'\n",
    "d_reg_num_name['17'] = 'Southern Andes'\n",
    "d_reg_num_name['18'] = 'New Zealand'\n",
    "d_reg_num_name['19'] = 'Antarctic and Subantarctic'\n",
    "\n",
    "d_reg_num_name_icci = {}\n",
    "d_reg_num_name_icci_reg = {'low-latitudes':['16'], 'mid-latitudes':['08', '11','02','18'], 'polar-hma':['06','13-14-15','01','04','17']}\n",
    "d_reg_num_name_icci_reg['Glaciers globally'] = ['global']\n",
    "\n",
    "d_reg_num_name_icci['global'] = 'Glaciers globally'\n",
    "d_reg_num_name_icci['01'] = 'Glaciers of Alaska'\n",
    "d_reg_num_name_icci['02'] = 'Glaciers of Western Canada & US'\n",
    "d_reg_num_name_icci['03'] = 'Glaciers of Arctic Canada (North)'\n",
    "d_reg_num_name_icci['04'] = 'Glaciers of Arctic Canada (South)'\n",
    "d_reg_num_name_icci['05'] = 'Glaciers of Greenland'\n",
    "d_reg_num_name_icci['06'] = 'Glaciers of Iceland'\n",
    "d_reg_num_name_icci['07'] = 'Glaciers of Svalbard and Jan Mayen'\n",
    "d_reg_num_name_icci['08'] = 'Glaciers of Scandinavia'\n",
    "d_reg_num_name_icci['09'] = 'Glaciers of Russian Arctic'\n",
    "d_reg_num_name_icci['10'] = 'Glaciers of North Asia'\n",
    "d_reg_num_name_icci['11'] = 'Glaciers of the Alps'\n",
    "d_reg_num_name_icci['12'] = 'Glaciers of Caucasus and Middle East'\n",
    "d_reg_num_name_icci['13'] = 'Glaciers of Central Asia'\n",
    "d_reg_num_name_icci['13-14-15'] = 'Glaciers of High Mountain Asia'\n",
    "d_reg_num_name_icci['14'] = 'Glaciers of South Asia West'\n",
    "d_reg_num_name_icci['15'] = 'Glaciers of South Asia East'\n",
    "d_reg_num_name_icci['16'] = 'Tropical Glaciers'\n",
    "d_reg_num_name_icci['17'] = 'Glaciers of Patagonia & Southern Andes'\n",
    "d_reg_num_name_icci['18'] = 'Glaciers of New Zealand'\n",
    "d_reg_num_name_icci['19'] = 'Glaciers of Antarctic and Subantarctic'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31eb577d",
   "metadata": {},
   "source": [
    "We have extracted the global climate change data in another notebook:\n",
    "- click on this link to see it: https://nbviewer.org/urls/cluster.klima.uni-bremen.de/~oggm/oggm-standard-projections/analysis_notebooks/global_gcm_climate_cmip6_cmip5.ipynb\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "5d48b4c1",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/lilianschuster/anaconda3/envs/oggm_v16/lib/python3.11/site-packages/pyproj/__init__.py:89: UserWarning: pyproj unable to set database path.\n",
      "  _pyproj_global_context_initialize()\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.427504187860568 6.274634111566688\n"
     ]
    }
   ],
   "source": [
    "   \n",
    "# load global mean temperature change files \n",
    "# this is how we extracted the global climate change data from:\n",
    "# https://cluster.klima.uni-bremen.de/~oggm/oggm-standard-projections/analysis_notebooks/global_gcm_climate_cmip6_cmip5.ipynb\n",
    "# note that the global mean temperature (GMT) changes have been computed by using the assumption that GMT changed from 1850-1900 to 1986-2005 by 0.63°C \n",
    "# (this is the assumption used by the IPCC SROCC -> --> https://www.ipcc.ch/srocc/chapter/summary-for-policymakers/\n",
    "# IPCC, 2019: Summary for Policymakers. In: IPCC Special Report on the Ocean and Cryosphere in a Changing Climate [H.-O. Pörtner, D.C. Roberts, V. Masson-Delmotte,\n",
    "# P. Zhai, M. Tignor, E. Poloczanska, K. Mintenbeck, A. Alegría, M. Nicolai, A. Okem, J. Petzold, B. Rama, N.M. Weyer (eds.)]. In press.\n",
    "# -> future follow-up studies should use the updated estimates (e.g. IPCC AR6 WG1 updated the methods and found a GMT change of 0.69°C from 1850-1900 and 1986-2005)\n",
    "try:\n",
    "    import oggm\n",
    "    dpath = 'https://cluster.klima.uni-bremen.de/~oggm/oggm-standard-projections/analysis_notebooks'\n",
    "    _file_cmip5 = oggm.utils.file_downloader(f'{dpath}/Global_mean_temp_deviation_2071_2100_2081_2100_2271_2300_2281_2300_rel_1850_1900_cmip5_gcms.csv',\n",
    "                                             reset=True)\n",
    "    _file_cmip6 = oggm.utils.file_downloader(f'{dpath}/Global_mean_temp_deviation_2071_2100_2081_2100_2271_2300_2281_2300_rel_1850_1900_cmip6_gcms.csv',\n",
    "                                             reset=True)\n",
    "except:\n",
    "    # you can also just download the file, \n",
    "    # e.g. via https://cluster.klima.uni-bremen.de/~oggm/oggm-standard-projections/analysis_notebooks/Global_mean_temp_deviation_2071_2100_2081_2100_2271_2300_2281_2300_rel_1850_1900_cmip5_gcms.csv\n",
    "    # and via https://cluster.klima.uni-bremen.de/~oggm/oggm-standard-projections/analysis_notebooks/Global_mean_temp_deviation_2071_2100_2081_2100_2271_2300_2281_2300_rel_1850_1900_cmip6_gcms.csv' \n",
    "    # change this path to your local path \n",
    "    _file_cmip5 = 'change_to_local_path'\n",
    "    _file_cmip6 = 'change_to_local_path'  \n",
    "    \n",
    "pd_cmip5_temp_ch_2100 = pd.read_csv(_file_cmip5, index_col=0)\n",
    "pd_cmip5_temp_ch_2100['cmip'] = 'CMIP5'\n",
    "pd_cmip6_temp_ch_2100 = pd.read_csv(_file_cmip6, index_col=0)\n",
    "pd_cmip6_temp_ch_2100['cmip'] = 'CMIP6'\n",
    "\n",
    "pd_cmip_temp_ch_2100 = pd.concat([pd_cmip6_temp_ch_2100, pd_cmip5_temp_ch_2100])\n",
    "pd_cmip_temp_ch_2100 = pd_cmip_temp_ch_2100.loc[pd_cmip_temp_ch_2100.ssp != 'ssp534-over']\n",
    "pd_cmip_temp_ch_2100_gcm_until_2300 = pd_cmip_temp_ch_2100.loc[pd_cmip_temp_ch_2100['global_temp_ch_2271-2300_preindustrial'].dropna().index]\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "data_temp = pd_cmip_temp_ch_2100_gcm_until_2300['global_temp_ch_2071-2100_preindustrial'].values.reshape(1,-1).T\n",
    "scaler = MinMaxScaler()\n",
    "scaler.fit(data_temp)\n",
    "t_min, t_max = pd_cmip_temp_ch_2100_gcm_until_2300['global_temp_ch_2071-2100_preindustrial'].min(), pd_cmip_temp_ch_2100_gcm_until_2300['global_temp_ch_2071-2100_preindustrial'].max()\n",
    "print(t_min, t_max)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "414e8215",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/html": [
       "<div style=\"vertical-align: middle;\"><strong>bins_100</strong> </div><div class=\"cmap\"><img alt=\"bins_100 colormap\" title=\"bins_100\" style=\"border: 1px solid #555;\" src=\"data:image/png;base64,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\"></div><div style=\"vertical-align: middle; max-width: 514px; display: flex; justify-content: space-between;\"><div style=\"float: left;\"><div title=\"#6495edff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #6495edff;\"></div> under</div><div style=\"margin: 0 auto; display: inline-block;\">bad <div title=\"#00000000\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #00000000;\"></div></div><div style=\"float: right;\">over <div title=\"#980002ff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #980002ff;\"></div></div>"
      ],
      "text/plain": [
       "<matplotlib.colors.LinearSegmentedColormap at 0x7f5ec8f9b5d0>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "######################\n",
    "\n",
    "from scipy.optimize import curve_fit\n",
    "\n",
    "# the exponential decay fit is just to compare to the LOWESS fit (but we only use it in the appendix_fit figures)\n",
    "def exponential_decay(x, a, b):\n",
    "    # avoind overflow -> does not work ... \n",
    "    exp_values = -b * x\n",
    "    #exp_values = np.clip(exp_values, -709, 709)  # Clip values to avoid overflow\n",
    "    return a * np.exp(exp_values) #+ c * np.exp(-d * x**2)#**c \n",
    "\n",
    "\n",
    "param_bounds = ([0, 0], [2000, 5])\n",
    "\n",
    "rgi_regs_global = ['global','01', '02', '03', '04', '05', '06', '07',\n",
    "                   '08', '09', '10', '11', '12', '13', '14', '15',\n",
    "                   '16', '17', '18', '19']\n",
    "rgi_regs_global.append('13-14-15')\n",
    "path_oggm = 'glacier_model_csv_files/OGGM/'\n",
    "path_glogem = 'glacier_model_csv_files/GloGEM/'\n",
    "path_pygem =  'glacier_model_csv_files/PyGEM-OGGM/'\n",
    "\n",
    "path_d = {}\n",
    "path_d['OGGM']  = path_oggm\n",
    "path_d['GloGEM'] = path_glogem\n",
    "path_d['PyGEM-OGGM'] = path_pygem\n",
    "\n",
    "# we sort it here into temperature levels, but we only use it internally to do the plots \n",
    "colors_temp_values = {'<=+1.5°C':1.5,'<=+2.0°C':2.0,'<=+3.0°C':3.0, '<=+4.0°C':4.0} #, '>4.0°C':5.0}\n",
    "colors_temp = {'<=+1.5°C':'cyan','<=+2.0°C':'blue','<=+3.0°C':'orange', '<=+4.0°C':'red', '>4.0°C':'purple'}\n",
    "\n",
    "from matplotlib.colors import LinearSegmentedColormap\n",
    "colors_temp4_l = ['cornflowerblue', 'gold', 'orange', 'darkorange', 'crimson', '#980002', '#980002', '#980002']\n",
    "cmap_blue_orange = LinearSegmentedColormap.from_list('bins_100',colors_temp4_l, N=1000)\n",
    "cmap_blue_orange"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "884be89b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# this just had to be done once, and can be skipped afterwards\n",
    "### Create all globally\n",
    "create_glogem_globally = False\n",
    "if create_glogem_globally:\n",
    "    # globally: sum over all RGI regions\n",
    "    from oggm.utils import mkdir\n",
    "\n",
    "    ssps_2300 = ['ssp126','ssp534-over', 'ssp585']\n",
    "    for var in ['volume', 'area']:\n",
    "        cmip == 'CMIP6' \n",
    "        end_years = [2300]\n",
    "        for end_year in end_years:\n",
    "            odir = f'{path_glogem}{var}/{cmip}/{end_year}/global/'\n",
    "            mkdir(odir)\n",
    "            for scenario in ssps_2300:\n",
    "                odf = 0\n",
    "                for n in np.arange(1,20,1):\n",
    "                    rgi_reg = f'RGI{n:02d}'\n",
    "                    idir = f'{path_glogem}{var}/{cmip}/{end_year}/{rgi_reg}/'\n",
    "                    df = pd.read_csv(idir + f'/{scenario}.csv', index_col=0)\n",
    "                    odf += df\n",
    "                odf.to_csv(odir + f'/{scenario}.csv')\n",
    "\n",
    "### Create all  models global_no_rgi05_rgi19\n",
    "create_global_no_rgi05_rgi19 = False\n",
    "if create_global_no_rgi05_rgi19:\n",
    "    # globally: sum over all RGI regions\n",
    "    from oggm.utils import mkdir\n",
    "\n",
    "    ssps_2300 = ['ssp126','ssp534-over', 'ssp585']\n",
    "    \n",
    "    for m in path_d.keys():\n",
    "        path_model = path_d[m]\n",
    "        for var in ['volume']:\n",
    "            cmip == 'CMIP6' \n",
    "            end_years = [2300]\n",
    "            for end_year in end_years:\n",
    "                odir = f'{path_model}{var}/{cmip}/{end_year}/global_no_rgi05_rgi19/'\n",
    "                mkdir(odir)\n",
    "                for scenario in ssps_2300:\n",
    "                    odf = 0\n",
    "                    for n in [ 1,  2,  3,  4,  #5,\n",
    "                              6,  7,  8,  9, 10, 11, 12, 13, 14, 15, 16, 17, 18, #19\n",
    "                              ]:\n",
    "                        rgi_reg = f'RGI{n:02d}'\n",
    "                        idir = f'{path_glogem}{var}/{cmip}/{end_year}/{rgi_reg}/'\n",
    "                        df = pd.read_csv(idir + f'/{scenario}.csv', index_col=0)\n",
    "                        odf += df\n",
    "                    odf.to_csv(odir + f'/{scenario}.csv')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "070020c8",
   "metadata": {},
   "source": [
    "**lowess fits with MoePy**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1a835ffb",
   "metadata": {},
   "outputs": [],
   "source": [
    "# this is just an algorithm to select the lowess \"frac\"- parameter,\n",
    "# which basiclly describes the fraction in-between which linearity is assumed\n",
    "# we use this algorithm together with some expert knowledge choices\n",
    "def select_frac_alg(sel_it, sim_year=2100):            \n",
    "    repeat = True\n",
    "    problem = False\n",
    "    diff_min = 0\n",
    "    min_add_q50 = 0\n",
    "    while repeat:\n",
    "        _sel = sel_it.loc[sel_it.year == sim_year].copy()\n",
    "        if sim_year == 2300:\n",
    "            # brute-force, rather try not to be completely linear for year 2300 \n",
    "            _sel = _sel.loc[_sel.frac<=0.75]\n",
    "\n",
    "        # also make sure to select a non-negative median\n",
    "        # this does NOT work in some regions, we have to clip negative values afterwards ... \n",
    "        _sel = _sel.loc[(_sel['min_q50']>= -1+min_add_q50) | (_sel['min_q50'] >= _sel['min_q50'].max())]\n",
    "\n",
    "        # first choose only those fracs that somehow are the most \"monotonic\" decreasing \n",
    "        min_q_diff = _sel[['min_q05_diff','min_q25_diff','min_q50_diff','min_q75_diff','min_q95_diff']].min(axis=1)\n",
    "\n",
    "        _sel.loc[min_q_diff.index,'min_q_diff'] = min_q_diff.values\n",
    "        max_min_q_diff = _sel['min_q_diff'].max()\n",
    "        # we add -0.2 to allow not only the best \"frac\", but also others ... \n",
    "        _sel = _sel.loc[_sel['min_q_diff'] >= max_min_q_diff-0.2]\n",
    "\n",
    "        _sel = _sel.loc[_sel.min_q50_diff>=-0.01-diff_min/5]\n",
    "        _sel = _sel.loc[_sel.min_q95_diff>=-0.01-diff_min]\n",
    "        _sel = _sel.loc[_sel.min_q75_diff>=-0.01-diff_min]\n",
    "        _sel = _sel.loc[_sel.min_q25_diff>=-0.01-diff_min]\n",
    "        _sel = _sel.loc[_sel.min_q05_diff>=-0.01-diff_min]\n",
    "\n",
    "        _val = _sel.min_q50_diff.max() if _sel.min_q50_diff.max() <=0 else 0\n",
    "        _sel = _sel.loc[_sel.min_q50_diff >=_val*1.01]\n",
    "\n",
    "        _val = _sel.min_q50.max() if _sel.min_q50.max() <=0 else 0\n",
    "        _sel = _sel.loc[_sel.min_q50 >=_val]\n",
    "        _val = _sel.min_q05.max() if _sel.min_q05.max()<=0 else 0\n",
    "        _sel = _sel.loc[_sel.min_q05 >=_val]\n",
    "        if len(_sel) == 97:\n",
    "            frac = np.round(_sel.frac.unique(),2)[0]\n",
    "            repeat = False\n",
    "\n",
    "        elif len(_sel)>97:\n",
    "            frac = np.round(_sel.frac.unique().max(),2)\n",
    "            _sel = _sel.loc[_sel.frac == _sel.frac.unique().max()]\n",
    "            repeat = False\n",
    "        else:\n",
    "            repeat = True\n",
    "            diff_min = diff_min + 0.01\n",
    "        if max_min_q_diff <-0.35:\n",
    "            repeat =True\n",
    "            min_add_q50 -= 0.5\n",
    "            if min_add_q50 < -5:\n",
    "                repeat = False\n",
    "                #frac = np.round(_sel.frac.unique().max(),2)\n",
    "                #_sel = _sel.loc[_sel.frac == _sel.frac.unique().max()]\n",
    "                problem = True\n",
    "\n",
    "    repeat = True\n",
    "    while problem:\n",
    "        print(rgi_reg, 'problem')\n",
    "        problem = False\n",
    "        diff_min = 0\n",
    "        min_add_q50 = 0\n",
    "        while repeat:\n",
    "            _sel = sel_it.loc[sel_it.year == sim_year].copy()\n",
    "            if sim_year == 2300:\n",
    "                # brute-force, ok, let's be a bit \"more\" linear\n",
    "                _sel = _sel.loc[_sel.frac<=0.85]\n",
    "\n",
    "            # also make sure to select a non-negative median\n",
    "            # this does NOT work in some regions, we have to clip negative values afterwards ... \n",
    "            _sel = _sel.loc[(_sel['min_q50']>= -2+min_add_q50) | (_sel['min_q50'] >= _sel['min_q50'].max())]\n",
    "\n",
    "            # first choose only those fracs that somehow are the most \"monotonic\" decreasing \n",
    "            min_q_diff = _sel[['min_q05_diff','min_q25_diff','min_q50_diff','min_q75_diff','min_q95_diff']].min(axis=1)\n",
    "\n",
    "            _sel.loc[min_q_diff.index,'min_q_diff'] = min_q_diff.values\n",
    "            max_min_q_diff = _sel['min_q_diff'].max()\n",
    "            # we add -0.2 to allow not only the best \"frac\", but also others ... \n",
    "            _sel = _sel.loc[_sel['min_q_diff'] >= max_min_q_diff-0.2]\n",
    "\n",
    "            _sel = _sel.loc[_sel.min_q50_diff>=-0.01-diff_min/5]\n",
    "            _sel = _sel.loc[_sel.min_q95_diff>=-0.01-diff_min]\n",
    "            _sel = _sel.loc[_sel.min_q75_diff>=-0.01-diff_min]\n",
    "            _sel = _sel.loc[_sel.min_q25_diff>=-0.01-diff_min]\n",
    "            _sel = _sel.loc[_sel.min_q05_diff>=-0.01-diff_min]\n",
    "\n",
    "            _val = _sel.min_q50_diff.max() if _sel.min_q50_diff.max() <=0 else 0\n",
    "            _sel = _sel.loc[_sel.min_q50_diff >=_val*1.01]\n",
    "\n",
    "            _val = _sel.min_q50.max() if _sel.min_q50.max() <=0 else 0\n",
    "            _sel = _sel.loc[_sel.min_q50 >=_val]\n",
    "            _val = _sel.min_q05.max() if _sel.min_q05.max()<=0 else 0\n",
    "            _sel = _sel.loc[_sel.min_q05 >=_val]\n",
    "            if len(_sel) == 97:\n",
    "                frac = np.round(_sel.frac.unique(),2)[0]\n",
    "                repeat = False\n",
    "\n",
    "            elif len(_sel)>97:\n",
    "                frac = np.round(_sel.frac.unique().max(),2)\n",
    "                _sel = _sel.loc[_sel.frac == _sel.frac.unique().max()]\n",
    "                repeat = False\n",
    "            else:\n",
    "                repeat = True\n",
    "                diff_min = diff_min + 0.02\n",
    "            if max_min_q_diff <-0.5:\n",
    "                repeat =True\n",
    "                min_add_q50 -= 0.5\n",
    "                if min_add_q50 < -5:\n",
    "                    repeat = False\n",
    "                    problem = True\n",
    "                    frac = np.NaN\n",
    "\n",
    "    print(rgi_reg, frac, min_add_q50, _sel['min_q50'].min(), max_min_q_diff)\n",
    "    return frac, _sel\n",
    "                "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a9b0eadd",
   "metadata": {},
   "source": [
    "**first check the frac dependency, we have applied some expert knowledge in some regions (after having checked the frac dependency)**\n",
    "- decided by Fabien Maussion and Lilian Schuster"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "53cc941f",
   "metadata": {},
   "outputs": [],
   "source": [
    "frac_2300_exp_decision = {'RGI03':0.95, 'RGI07':0.95, 'RGI09':0.8,   # RGI13 0.8 ok!\n",
    "                          'RGI12':0.63, 'RGI14':0.56, 'RGI15':0.5, 'RGI19':0.59,\n",
    "                          'RGI17':0.8, 'RGI18':0.8,'RGI02':0.8,  \n",
    "                          'RGI16':0.7, 'RGI11':0.6, 'RGI08':0.7}\n",
    "\n",
    "def select_frac_w_experts(sel_it, sim_year=2300):\n",
    "    region = sel_it.region.unique()[0]\n",
    "    if (region in frac_2300_exp_decision.keys()) and (sim_year == 2300):\n",
    "        frac = frac_2300_exp_decision[region]\n",
    "        _sel = sel_it.loc[sel_it.year == sim_year].copy()\n",
    "        _sel = _sel.loc[_sel.frac.round(2) == frac]\n",
    "    else:\n",
    "        frac, _sel = select_frac_alg(sel_it, sim_year=sim_year)\n",
    "    return frac, _sel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "01c70bb4",
   "metadata": {},
   "outputs": [],
   "source": [
    "it = 1\n",
    "pd_fit_stats_it = pd.read_csv(f'lowess_fit_stats_oct29_predi_all_it{it}.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "570b6586",
   "metadata": {},
   "outputs": [],
   "source": [
    "pd_fit_stats_reg = pd_fit_stats_it.loc[pd_fit_stats_it.region == rgi_reg]\n",
    "pd_fit_stats_reg_lowess = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'lowess_predi']\n",
    "pd_fit_stats_reg_exp = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'exp']\n",
    "sel_it = pd_fit_stats_reg_lowess\n",
    "frac, _sel = select_frac_w_experts(sel_it, sim_year=sim_year)\n",
    "_sel[['region','year', 'deltaTemp', 'q50', 'q05','q25','q75','q95']].reset_index(drop=True).to_csv(f'lowess_fit_from_GCMs_until_2300_test_{rgi_reg}_{sim_year}.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "740f1d82",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI16 problem\n",
      "RGI16 0.8 -2.0 -3.6536700397631856 -0.3792888174594004\n",
      "RGI08 0.67 -0.5 -1.1500212197060011 -0.3084628888246981\n",
      "RGI11 problem\n",
      "RGI11 0.8 0 -1.8135322844575616 -0.2882249343312313\n",
      "RGI02 0.74 -3.5 -4.465209266971082 -0.3436600552033668\n",
      "RGI18 problem\n",
      "RGI18 0.8 -0.5 -2.4939965189887108 -0.4151130332801664\n",
      "RGI06 problem\n",
      "RGI06 0.8 0 -1.6090488460476096 -0.2584169616942668\n",
      "RGI13-14-15 0.57 0 -0.0536211913880723 -0.3423459518639633\n",
      "RGI01 0.57 0 0.4378871108982666 -0.0461755604138239\n",
      "RGI04 0.72 -0.5 -1.453974934319801 -0.1119918149230363\n",
      "RGI17 problem\n",
      "RGI17 0.43 0 0.0086394260179012 -0.3464151442898471\n",
      "global 0.45 0 0.4881256712591804 -0.18261132766947\n",
      "RGI03 problem\n",
      "RGI03 problem\n",
      "RGI03 nan 0 nan -0.5620769679095172\n",
      "RGI05 0.64 0 -0.6349166509171063 -0.3024389572831554\n",
      "RGI07 problem\n",
      "RGI07 0.37 0 -0.5285588061270116 -0.3772247901744023\n",
      "RGI09 0.4 0 0.0014548089734972 -0.1512704650111835\n",
      "RGI10 0.52 0 -0.6288899289452097 -0.2626732102131575\n",
      "RGI12 problem\n",
      "RGI12 0.35 0 -0.7280537981700164 -0.4506375881144771\n",
      "RGI13 problem\n",
      "RGI13 0.8 -1.5 -3.487859736807667 -0.290466212611608\n",
      "RGI14 0.6 0 0.003148286417572 -0.1897086239623608\n",
      "RGI15 0.6 0 -0.013475623951589 -0.2381557071270472\n",
      "RGI19 0.49 0 0.4480962529104073 -0.2585584699348254\n"
     ]
    }
   ],
   "source": [
    "\n",
    "d_reg_num_name_icci_reg['appendix'] = ['03','05','07','09','10','12','13','14','15','19']\n",
    "# from notebook: lowess_percentile_interval_fit_create_fit_data_and_merge_lowess_final_fit_files.ipynb\n",
    "pd_fit_data_all = pd.read_csv(f'fit_data_temp_evol_global_regional_oggm_pygem_glogem_oct182023.csv', index_col=0)\n",
    "sim_year = 2300\n",
    "for reg in list(d_reg_num_name_icci_reg.keys()):\n",
    "    rgi_regs = d_reg_num_name_icci_reg[reg]\n",
    "    #fig, axs = plt.subplots(3,2, figsize=(25,22)) #harey=True, sharex=True)\n",
    "    #fig_fit, axs_fit = plt.subplots(3,2, figsize=(25,22)) #harey=True, sharex=True)\n",
    "    #axs_fit = axs_fit.flatten()\n",
    "    #axs = axs.flatten()\n",
    "    len_sub_p = len(rgi_regs)\n",
    "\n",
    "    for jj,_rgi_reg in enumerate(rgi_regs): \n",
    "        list_index_l = []\n",
    "        err_l = []\n",
    "        for m in path_d.keys():\n",
    "            list_index_l.append(list(pd_cmip_temp_ch_2100.index + '_' + m))\n",
    "        pd_proj_2100_2300 = pd.DataFrame(index=np.concatenate(list_index_l))\n",
    "        #pd_proj_2100_2300 = pd.DataFrame(index=pd_cmip_temp_ch_2100.index)\n",
    "\n",
    "        if _rgi_reg !='global':\n",
    "            rgi_reg = f'RGI{_rgi_reg}'\n",
    "        else:\n",
    "            rgi_reg = _rgi_reg \n",
    "\n",
    "        data = pd_fit_data_all.loc[(pd_fit_data_all.year==sim_year) & (pd_fit_data_all.region==rgi_reg)]\n",
    "        # Generate data looking like cosine\n",
    "        x = data['global_temp_ch_2071-2100_preindustrial'].values\n",
    "        y = data['rel_ice_%_2020'].values        \n",
    "\n",
    "        pd_fit_stats_reg = pd_fit_stats_it.loc[pd_fit_stats_it.region == rgi_reg]\n",
    "        pd_fit_stats_reg_lowess = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'lowess_predi']\n",
    "        pd_fit_stats_reg_exp = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'exp']\n",
    "        sel_it = pd_fit_stats_reg_lowess # we don't have it anymore ... .loc[pd_fit_stats_reg_lowess.it ==it]\n",
    "        frac_alg, _sel_alg = select_frac_alg(sel_it, sim_year=sim_year)\n",
    "        frac_alg = np.round(frac_alg,2)\n",
    "        _sel = sel_it.loc[sel_it.year == sim_year].copy()\n",
    "        fracs = _sel.frac.unique()\n",
    "\n",
    "        fig, axs = plt.subplots(7,10, figsize=(70,50), sharey=True)\n",
    "        axs = axs.flatten()\n",
    "        jj=0\n",
    "        plt.suptitle(rgi_reg+ f', it={it}')\n",
    "        for j,frac in enumerate(fracs):\n",
    "            frac_r = np.round(frac,2)\n",
    "            _sel_f = _sel.loc[_sel.frac == frac].copy()\n",
    "            assert len(_sel_f.frac.unique()) == 1\n",
    "            min_qs = np.round(_sel_f[['min_q50','min_q05','min_q25','min_q75','min_q95']].min(axis=1).min(),2)\n",
    "            min_qs_diff = np.round(_sel_f[['min_q05_diff','min_q25_diff','min_q50_diff','min_q75_diff','min_q95_diff']].min(axis=1).min(),2)\n",
    "            ax = axs[jj]\n",
    "            ax.set_title(f'{np.round(frac,2)}')\n",
    "\n",
    "            ax.plot(_sel_f.deltaTemp,\n",
    "                     _sel_f.q50, lw=3, color='black', alpha = 1, label=f'lowess, frac={frac}')\n",
    "            ax.fill_between(_sel_f.deltaTemp, _sel_f.q05, _sel_f.q95, lw=3, alpha=0.2, color=f'black', label='90% PI')\n",
    "            ax.fill_between(_sel_f.deltaTemp, _sel_f.q25, _sel_f.q75, lw=3, alpha=0.4, color=f'black', label = '50% PI (IQR)')\n",
    "            ax.scatter(x,y, color = 'blue', label='model projections', s=40)\n",
    "            jj+=1\n",
    "            if rgi_reg == 'RGI08':\n",
    "                ax.set_ylim([-1,120])\n",
    "            ax.text(2.5, ax.get_ylim()[-1]*0.9,\n",
    "                                   f' min qs: {min_qs},\\nmin_qs_diff: {min_qs_diff}', fontsize=18)\n",
    "            \n",
    "            if (min_qs>=-3.5) and (min_qs_diff>=-0.9) and (_sel_f['min_q50_diff'].min()>=-0.2):\n",
    "                ax.text(4, ax.get_ylim()[-1]*0.7,\n",
    "                                   f' good fit\\nby algorithm 1', fontsize=22)\n",
    "            frac_r == np.round(frac,2)\n",
    "            if frac_r == frac_alg:\n",
    "                ax.text(4, ax.get_ylim()[-1]*0.5,\n",
    "                                   f' chosen fit\\nby algorithm 2', fontsize=22, color='black')\n",
    "                if rgi_reg not in frac_2300_exp_decision.keys():\n",
    "                    ax.text(4, ax.get_ylim()[-1]*0.3,\n",
    "                                       f'final decision\\nby experts', fontsize=28, color='red')\n",
    "            if rgi_reg in frac_2300_exp_decision.keys():\n",
    "                frac_exp = frac_2300_exp_decision[rgi_reg]\n",
    "                if frac_r == frac_exp:\n",
    "                    ax.text(4, ax.get_ylim()[-1]*0.3,\n",
    "                                       f'final decision\\nby experts', fontsize=28, color='red')\n",
    "           \n",
    "        plt.tight_layout()\n",
    "        fig.savefig(f'figures/appendix_fracs/{sim_year}_lowess_fracs_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3_{it}.png')\n",
    "        plt.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "982310bc",
   "metadata": {},
   "source": [
    "# NOW the actual plots:\n",
    "**without recomputing the lowess values by just showing the deciles**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "daa8971b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI16 0.41 0 0.5940125822019444 -0.009331074123102\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI08 0.96 0 1.0382209879829225 0.0437381748806752\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI11 0.51 0 0.4329462619194236 -0.0115070402162595\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI02 0.46 0 0.0863687406451393 -0.0073848595658214\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI18 0.86 0 2.421712130179934 0.0639439491791801\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI06 0.99 0 21.31067898391548 0.075307676741005\n",
      "RGI06 problem\n",
      "RGI06 0.8 0 -1.6090488460476096 -0.2584169616942668\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13-14-15 0.99 0 8.45265502846896 0.0300477867204858\n",
      "RGI13-14-15 0.57 0 -0.0536211913880723 -0.3423459518639633\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI01 0.99 0 24.468128333237928 0.0562669460517852\n",
      "RGI01 0.57 0 0.4378871108982666 -0.0461755604138239\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI04 0.97 0 17.45129269512268 -0.2584482711844496\n",
      "RGI04 0.72 -0.5 -1.453974934319801 -0.1119918149230363\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI17 0.99 0 16.27953236489871 0.3190121418276295\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "global 0.99 0 40.35323776152246 0.1281386443393088\n",
      "global 0.45 0 0.4881256712591804 -0.18261132766947\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI03 problem\n",
      "RGI03 0.86 0 65.2129171468793 -0.379642069470087\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI05 0.99 0 31.33411392714501 0.0360290568379184\n",
      "RGI05 0.64 0 -0.6349166509171063 -0.3024389572831554\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI07 0.76 0 11.748008008139507 -0.0234831019352412\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI09 0.99 0 32.842798576655184 -0.1408007488342377\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI10 0.99 0 1.6764143761648478 0.0695534992579747\n",
      "RGI10 0.52 0 -0.6288899289452097 -0.2626732102131575\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI12 0.7 0 0.8786420268992139 0.0240430667407689\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13 0.99 0 3.4005703198600625 0.0456394446257517\n",
      "RGI13 problem\n",
      "RGI13 0.8 -1.5 -3.487859736807667 -0.290466212611608\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI14 0.99 0 16.40754998175739 0.0381533680331376\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI15 0.78 0 1.3234315964212937 0.0325894930033283\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI19 0.99 0 61.57731305368136 0.0930242333842912\n"
     ]
    }
   ],
   "source": [
    "# this file has been merged in lowess_percentile_interval_fit_create_fit_data_and_merge_lowess_final_fit_files.ipynb\n",
    "# the actual fits were computed by lowess_percentile_interval_fit_per_region.py\n",
    "pd_fit_stats = pd.read_csv('lowess_fit_stats_oct29_predi_all_it1.csv', index_col=0)\n",
    "def boxplot_manual(qs= [0.1,0.5,1,2,3],\n",
    "                   p = 2300, ax = None, j=0, temp_val = 1.5):\n",
    "    q05, q25, q50, q75, q95 = qs\n",
    "    c = cmap(scaler.transform(np.array([temp_val]).reshape(-1,1))).flatten()\n",
    "    item = {}\n",
    "    #item[\"label\"] = 'box' # not required\n",
    "    item[\"med\"] = q50\n",
    "    item[\"q1\"] = q25\n",
    "    item[\"q3\"] = q75\n",
    "    #item[\"cilo\"] = 5.3 # not required\n",
    "    #item[\"cihi\"] = 5.7 # not required\n",
    "    item[\"whislo\"] = q05\n",
    "    item[\"whishi\"] = q95\n",
    "    stats = [item]\n",
    "    boxprops = dict(linestyle='-', linewidth=2, color=c)\n",
    "    whisprops = dict(color=c, linewidth=1)\n",
    "    flierprops = dict(marker='o', markerfacecolor='green', markersize=0,\n",
    "                      markeredgecolor='none')\n",
    "    medianprops = dict(linestyle='-', linewidth=4, color=c)\n",
    "    ax.bxp(stats, positions = [p + j], showfliers = False,\n",
    "            flierprops= flierprops, medianprops=medianprops, boxprops=boxprops,\n",
    "             widths=3, #patch_artist=True,\n",
    "            whiskerprops=whisprops, capprops = dict(color=c, linewidth=1))\n",
    "    \n",
    "add_median = True\n",
    "color_continuous = True\n",
    "decile_ticks = True\n",
    "plt.rc('font', size=26)\n",
    "it = 1\n",
    "col_opt = 'blue_orange'\n",
    "frac_dict = {}\n",
    "for uncertainty in ['boxplot']: #, 'no_uncertainties']: # 'no_uncertainties',\n",
    "    if uncertainty == 'boxplot':\n",
    "        num_fits = 1000\n",
    "        delta_text = 20\n",
    "        xlim1 = 2325\n",
    "    else:\n",
    "        num_fits = 1000 \n",
    "        delta_text = 4\n",
    "        xlim1 = 2309\n",
    "    if col_opt =='cyan':\n",
    "        cmap = cmap_cyan\n",
    "    elif col_opt == 'blue':\n",
    "        cmap = cmap_blue\n",
    "    elif col_opt == 'blue_orange':\n",
    "        cmap = cmap_blue_orange\n",
    "    for reg in list(d_reg_num_name_icci_reg.keys()):\n",
    "        rgi_regs = d_reg_num_name_icci_reg[reg]\n",
    "        \n",
    "        len_sub_p = len(rgi_regs)\n",
    "\n",
    "        for jj,_rgi_reg in enumerate(rgi_regs): \n",
    "            list_index_l = []\n",
    "            err_l = []\n",
    "            for m in path_d.keys():\n",
    "                list_index_l.append(list(pd_cmip_temp_ch_2100.index + '_' + m))\n",
    "            pd_proj_2100_2300 = pd.DataFrame(index=np.concatenate(list_index_l))\n",
    "\n",
    "            if _rgi_reg !='global':\n",
    "                rgi_reg = f'RGI{_rgi_reg}'\n",
    "            else:\n",
    "                rgi_reg = _rgi_reg \n",
    "                \n",
    "            ####\n",
    "            pd_fit_stats_reg = pd_fit_stats.loc[pd_fit_stats.region == rgi_reg]\n",
    "            pd_fit_stats_reg_lowess = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'lowess_predi']\n",
    "            pd_fit_stats_reg_exp = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'exp']\n",
    "            sel_it = pd_fit_stats_reg_lowess # we don't have it anymore ... .loc[pd_fit_stats_reg_lowess.it ==it]\n",
    "\n",
    "            ######\n",
    "            fig, ax = plt.subplots(1,1, figsize=(15,10)) \n",
    "            fig_fit, axs_fit = plt.subplots(2,1, figsize=(15,20)) \n",
    "            \n",
    "            for end_year,ls in zip(['2300'],['-']):\n",
    "                #for temp_lev in colors_temp.keys():\n",
    "\n",
    "                for cmip in ['CMIP5','CMIP6']:\n",
    "                    pd_cmip = pd_cmip_temp_ch_2100_gcm_until_2300.loc[pd_cmip_temp_ch_2100_gcm_until_2300.ssp != 'ssp534-over']\n",
    "                    pd_cmip = pd_cmip.loc[pd_cmip.cmip == cmip]\n",
    "                    # check that there are temperature values in 2300 available ... \n",
    "                    pd_cmip = pd_cmip.loc[pd_cmip['global_temp_ch_2271-2300_preindustrial'].dropna().index]\n",
    "                    for gcm_scenario in pd_cmip.index:\n",
    "                        df_rel_l = []\n",
    "                        for m in path_d.keys():\n",
    "                            try:\n",
    "                                gcm, scenario = gcm_scenario.split('_')\n",
    "                                path = path_d[m]\n",
    "                                gcm_scenario_m = gcm_scenario + '_' + m\n",
    "                                n = len(gcm_scenario)\n",
    "                                if _rgi_reg == '13-14-15':\n",
    "                                    df = 0\n",
    "                                    for sub_rgi_reg in ['RGI13','RGI14','RGI15']:\n",
    "                                        idir = path + f'volume/{cmip}/{end_year}/{sub_rgi_reg}/'\n",
    "                                        _df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                        _df.columns = [g.upper() for g in _df.columns]\n",
    "                                        _df = _df[gcm.upper()]\n",
    "                                        df += _df\n",
    "                                else:\n",
    "                                    idir = path  + f'volume/{cmip}/{end_year}/{rgi_reg}/'\n",
    "                                    df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                    df.columns = [g.upper() for g in df.columns]\n",
    "                                    df = df[gcm.upper()]\n",
    "                                df = df.loc[2000:]\n",
    "                                # first do the rolling than the relative estimates .. \n",
    "                                df_roll = df.rolling(20, center=True, min_periods=1).mean().dropna()\n",
    "                                df_rel_roll = 100*df_roll/df_roll.loc[2020]                                    \n",
    "\n",
    "                                temp_ch = pd_cmip.loc[gcm_scenario, 'global_temp_ch_2071-2100_preindustrial']\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, 'global_temp_ch_2071-2100_preindustrial'] = temp_ch\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2100_{rgi_reg}'] = df_rel_roll.loc[2100]\n",
    "\n",
    "                                color= cmap(scaler.transform(temp_ch.reshape(-1,1))).flatten()\n",
    "                                if not add_median:\n",
    "                                    alpha = 0.2\n",
    "                                    ax.plot(df_rel_roll.loc[2020:2300].index, df_rel_roll.loc[2020:2300], lw=2, color=color,ls=ls, alpha = alpha) \n",
    "                                    ax.plot(df_rel_roll.loc[2000:2020].index, df_rel_roll.loc[2000:2020], lw=3, color='black', ls=ls, alpha = alpha)\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2300_{rgi_reg}'] = df_rel_roll.loc[2300]\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, 'model'] = m\n",
    "\n",
    "                                df_rel_l.append(df_rel_roll.loc[2000:2300])\n",
    "                            except:\n",
    "                                pass\n",
    "                        if add_median:\n",
    "                            if len(df_rel_l) > 0:\n",
    "                                df_rel_med = pd.DataFrame(df_rel_l).T.median(axis=1)\n",
    "                                ax.plot(df_rel_med.loc[2020:2300].index, df_rel_med.loc[2020:2300], lw=2, color=color,ls=ls, alpha = 0.4) \n",
    "                                ax.plot(df_rel_med.loc[2000:2020].index, df_rel_med.loc[2000:2020], lw=2, color='black', ls=ls, alpha = 0.2)\n",
    "\n",
    "            ax.set_ylim([0,125])\n",
    "            ax.set_ylabel('% of Remaining Ice',# (relative to 2020)',\n",
    "                          fontweight = 'bold')\n",
    "            ax.set_xlabel('Year',  fontweight = 'bold')\n",
    "            handles, labels = ax.get_legend_handles_labels()\n",
    "            for h in handles:\n",
    "                h.set_linewidth(3)\n",
    "            ax.legend(handles, labels,\n",
    "                          handlelength=1,\n",
    "                          title='Temperature change between 1850-1900 & 2071-2100:', ncol=5)\n",
    "            ax.legend().remove()\n",
    "\n",
    "            ax.set_title(d_reg_num_name_icci[_rgi_reg], fontweight='bold')\n",
    "\n",
    "            pd_proj_2100_2300 = pd_proj_2100_2300.dropna()\n",
    "            assert len(pd_proj_2100_2300) == 60\n",
    "            for j,sim_year in enumerate([2100,2300]):\n",
    "                # choose a \"frac\" parameter and select the right fitting values \n",
    "                frac, _sel = select_frac_w_experts(sel_it, sim_year=sim_year)\n",
    "                frac_dict[f'{rgi_reg}_{sim_year}'] = frac\n",
    "                \n",
    "                ax_fit = axs_fit[j]\n",
    "                ax_fit.set_title(f'{d_reg_num_name_icci[_rgi_reg]}:  year {sim_year}', fontweight='bold')\n",
    "\n",
    "                # get the raw data ...\n",
    "                x = pd_proj_2100_2300['global_temp_ch_2071-2100_preindustrial']\n",
    "                y = pd_proj_2100_2300[f'rel_ice_{sim_year}_{rgi_reg}']\n",
    "                \n",
    "                t_min, t_max = x.min(), x.max()\n",
    "                if sim_year == 2100:\n",
    "                    delta_x = 0.002\n",
    "                else:\n",
    "                    delta_x = 0.001\n",
    "\n",
    "                df_quantiles = _sel[['q50','q05','q25','q75','q95']]\n",
    "                # we just have to clip negative values .... there are still some although we tried to omit them by choosing the best frac ... \n",
    "                df_quantiles = df_quantiles.where(df_quantiles>=0, 0)\n",
    "                _sel_ice_eval = df_quantiles['q50'].values\n",
    "                _sel_ice_eval_bottom = df_quantiles['q05'].values\n",
    "                _sel_ice_eval_bottom_iqr = df_quantiles['q25'].values\n",
    "                _sel_ice_eval_top_iqr = df_quantiles['q75'].values\n",
    "                _sel_ice_eval_top = df_quantiles['q95'].values\n",
    "\n",
    "                ax_fit.plot(pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].deltaTemp,\n",
    "                         pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].q50, lw=3, color='green', alpha = 0.5, label='exponential fit')\n",
    "                ax_fit.plot(_sel.deltaTemp,\n",
    "                         _sel_ice_eval, lw=3, color='black', alpha = 1, label=f'lowess, frac={frac}')\n",
    "                ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom, _sel_ice_eval_top, lw=3, alpha=0.2, color=f'black', label='90% PI')\n",
    "                ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom_iqr, _sel_ice_eval_top_iqr, lw=3, alpha=0.4, color=f'black', label = '50% PI (IQR)')\n",
    "                \n",
    "                temp_eval_x = _sel.deltaTemp.values\n",
    "                sns.scatterplot(ax=ax_fit,\n",
    "                                data = pd_proj_2100_2300, #.loc[j], \n",
    "                                x='global_temp_ch_2071-2100_preindustrial', y=f'rel_ice_{sim_year}_{rgi_reg}',\n",
    "                                                style='model',\n",
    "                                hue='global_temp_ch_2071-2100_preindustrial',\n",
    "                                palette=cmap,      \n",
    "                                sizes=[150,150,150], size='model') #, legend=)\n",
    "                ax_fit.legend(fontsize=16, ncols=2)\n",
    "\n",
    "                ax_fit.set_xlabel('Temperature change between 1850-1900 & 2071-2100')\n",
    "                ax_fit.set_ylabel(f'% of Remaining Glacier Ice\\nin {sim_year} (relative to 2020)')\n",
    "\n",
    "                for j,temp_lev in enumerate(colors_temp.keys()):\n",
    "\n",
    "                    # for the annotation in the main plot ... \n",
    "                    if temp_lev != '>4.0°C':\n",
    "\n",
    "                        temp_lev_value = colors_temp_values[temp_lev]\n",
    "                        color= cmap(scaler.transform(np.array([temp_lev_value]).reshape(-1,1))).flatten()\n",
    "                        idx_nearest = np.abs(temp_eval_x - temp_lev_value).argmin()\n",
    "                        if j >0:\n",
    "                            # check the previous fit, if they overlap, need to shift the text\n",
    "                            previous_fit = fit.copy()\n",
    "                        else:\n",
    "                            previous_fit = 10\n",
    "                            \n",
    "                            \n",
    "                        if previous_fit <4:\n",
    "                            shift = previous_fit-3.5\n",
    "                        else:\n",
    "                            shift = 0\n",
    "                        \n",
    "                        fit = _sel_ice_eval[idx_nearest]\n",
    "\n",
    "\n",
    "                        if sim_year == 2300:\n",
    "                            loc_2300_stats = 2300 +4 \n",
    "                        else:\n",
    "                            loc_2300_stats = 2100\n",
    "                        #if temp_lev_value>1.5:\n",
    "                        #    ax.vlines(loc_2300_stats,y_fitted[0], y_fitted[1], color=colors_temp[temp_lev], lw=15, alpha = 0.4)\n",
    "                        ax.hlines(fit,loc_2300_stats-3, loc_2300_stats+3, lw=8, color=color, zorder=10)\n",
    "\n",
    "                        if sim_year == 2300:\n",
    "                            bbox=dict(facecolor='white', edgecolor='white', boxstyle='round', alpha = 0.8)\n",
    "                            ax.annotate(temp_lev[2:], xy=(2300, fit),\n",
    "                                        xytext=(loc_2300_stats+delta_text, fit+shift),\n",
    "                                                    verticalalignment=\"center\",\n",
    "                                                    color=color, alpha=1, fontweight = 'bold', fontsize=18)\n",
    "                            _sel_temp_lev_qs = np.array([ _sel_ice_eval_bottom[idx_nearest],\n",
    "                                                         _sel_ice_eval_bottom_iqr[idx_nearest], _sel_ice_eval[idx_nearest],\n",
    "                                                         _sel_ice_eval_top_iqr[idx_nearest], _sel_ice_eval_top[idx_nearest]])\n",
    "                            if uncertainty == 'boxplot':\n",
    "                                boxplot_manual(qs= _sel_temp_lev_qs,\n",
    "                                               p = loc_2300_stats+5.5,\n",
    "                                               ax = ax, j=j*4.1, temp_val = temp_lev_value,\n",
    "                                               )\n",
    "                np.testing.assert_allclose(temp_eval_x[::2], np.arange(1.4,6.3,0.1))\n",
    "                colors_fit = cmap(scaler.transform(temp_eval_x[::2].reshape(-1,1))) # only every decile ... \n",
    "                lw = 2\n",
    "                for c,f in zip(colors_fit, _sel_ice_eval[::2]):\n",
    "                    ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                               color=c, zorder=7)\n",
    "\n",
    "            ax.set_xlim([1997, xlim1])\n",
    "            if rgi_reg == 'RGI08':\n",
    "                ax_fit.set_ylim([-0.5, 120])\n",
    "\n",
    "            ax.set_xticks([2020,2100,2200,2300],[2020,2100,2200,2300])\n",
    "            ax.spines['right'].set_visible(False)\n",
    "            ax.spines['top'].set_visible(False)\n",
    "            fig.tight_layout()\n",
    "            fig_fit.tight_layout()\n",
    "            if uncertainty== 'boxplot':\n",
    "                \n",
    "                if add_median:\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/png/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3_{uncertainty}.png')\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/svg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3_{uncertainty}.svg')\n",
    "                    if uncertainty == 'boxplot':\n",
    "                        fig_fit.savefig(f'figures/appendix_fit/lowess_predi_exp_fit_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3.png')\n",
    "                        fig_fit.savefig(f'figures/appendix_fit/lowess_predi_exp_fit_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3.svg')\n",
    "                else:\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/png/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3_{uncertainty}.png')\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/svg/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3_{uncertainty}.svg')\n",
    "                    if uncertainty == 'boxplot':\n",
    "                        fig_fit.savefig(f'figures/appendix_fit/lowess_predi_exp_fit_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3.png')\n",
    "                        fig_fit.savefig(f'figures/appendix_fit/lowess_predi_exp_fit_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3.svg')\n",
    "\n",
    "            else:\n",
    "                if add_median:\n",
    "                    fig.savefig(f'figures/main_figures_without_uncertainties/median_glac_models/png/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3_{uncertainty}.png')\n",
    "                    fig.savefig(f'figures/main_figures_without_uncertainties/median_glac_models/svg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3_{uncertainty}.svg')\n",
    "                else:\n",
    "                    fig.savefig(f'figures/main_figures_without_uncertainties/all_glac_models/png/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3_{uncertainty}.png')\n",
    "                    fig.savefig(f'figures/main_figures_without_uncertainties/all_glac_models/svg/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3_{uncertainty}.svg')\n",
    "                    \n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3e07d4fb",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "3ebe03fa",
   "metadata": {},
   "source": [
    "**Create the same figures but now as subplots**\n",
    "-> actually, we decided to only show three temperature levels for the ICCI State of the Cryosphere report, and also did some other changes, so we leave this here just in case we want to go back ... "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "id": "387d444c",
   "metadata": {},
   "outputs": [],
   "source": [
    "show_all_four_labels = False\n",
    "if show_all_four_labels:\n",
    "\n",
    "    # this file has been merged in lowess_percentile_interval_fit_create_fit_data_and_merge_lowess_final_fit_files.ipynb\n",
    "    # the actual fits were computed by lowess_percentile_interval_fit_per_region.py\n",
    "    pd_fit_stats = pd.read_csv('lowess_fit_stats_oct29_predi_all_it1.csv', index_col=0)\n",
    "    def boxplot_manual(qs= [0.1,0.5,1,2,3],\n",
    "                       p = 2300, ax = None, j=0, temp_val = 1.5):\n",
    "        q05, q25, q50, q75, q95 = qs\n",
    "        c = cmap(scaler.transform(np.array([temp_val]).reshape(-1,1))).flatten()\n",
    "        item = {}\n",
    "        #item[\"label\"] = 'box' # not required\n",
    "        item[\"med\"] = q50\n",
    "        item[\"q1\"] = q25\n",
    "        item[\"q3\"] = q75\n",
    "        #item[\"cilo\"] = 5.3 # not required\n",
    "        #item[\"cihi\"] = 5.7 # not required\n",
    "        item[\"whislo\"] = q05\n",
    "        item[\"whishi\"] = q95\n",
    "        stats = [item]\n",
    "        boxprops = dict(linestyle='-', linewidth=2, color=c)\n",
    "        whisprops = dict(color=c, linewidth=1)\n",
    "        flierprops = dict(marker='o', markerfacecolor='green', markersize=0,\n",
    "                          markeredgecolor='none')\n",
    "        medianprops = dict(linestyle='-', linewidth=4, color=c)\n",
    "        ax.bxp(stats, positions = [p + j], showfliers = False,\n",
    "                flierprops= flierprops, medianprops=medianprops, boxprops=boxprops,\n",
    "                 widths=3, #patch_artist=True,\n",
    "                whiskerprops=whisprops, capprops = dict(color=c, linewidth=1))\n",
    "\n",
    "    add_median = True\n",
    "    color_continuous = True\n",
    "    decile_ticks = True\n",
    "    plt.rc('font', size=26)\n",
    "    it = 1\n",
    "    col_opt = 'blue_orange'\n",
    "    frac_dict = {}\n",
    "    for uncertainty in [ 'boxplot', 'no_uncertainties']: # 'no_uncertainties',\n",
    "        if uncertainty == 'boxplot':\n",
    "            num_fits = 1000\n",
    "            delta_text = 20\n",
    "            xlim1 = 2325\n",
    "        else:\n",
    "            num_fits = 1000 \n",
    "            delta_text = 4\n",
    "            xlim1 = 2309\n",
    "        if col_opt =='cyan':\n",
    "            cmap = cmap_cyan\n",
    "        elif col_opt == 'blue':\n",
    "            cmap = cmap_blue\n",
    "        elif col_opt == 'blue_orange':\n",
    "            cmap = cmap_blue_orange\n",
    "        for reg in list(d_reg_num_name_icci_reg.keys()):\n",
    "            rgi_regs = d_reg_num_name_icci_reg[reg]\n",
    "\n",
    "            len_sub_p = len(rgi_regs)\n",
    "            props = dict(boxstyle='round', facecolor='white', alpha=0.5)\n",
    "            if (reg == 'Glaciers globally') or (reg == 'low-latitudes'):\n",
    "                fig, axs = plt.subplots(1,1, figsize=(12.5,25/3)) \n",
    "                axs = [axs]#.flatten()\n",
    "                #for a in [1,2,3,4,5]:\n",
    "                #    axs[a].remove()\n",
    "                # figure to show caption \n",
    "                ax_f = axs[0] \n",
    "\n",
    "            elif reg == 'mid-latitudes':\n",
    "                fig, axs = plt.subplots(2,2, figsize=(25,25*2/3))\n",
    "                axs = axs.flatten()\n",
    "                #axs[4].remove()\n",
    "                #axs[5].remove()\n",
    "                ax_f = axs[2]\n",
    "            elif reg == 'polar-hma':\n",
    "                fig, axs = plt.subplots(3,2, figsize=(25,25))\n",
    "                axs = axs.flatten()\n",
    "                ax_f = axs[4]\n",
    "                #ax_f.spines['left'].set_visible(False)\n",
    "                #ax_f.spines['bottom'].set_visible(False)\n",
    "                axs[5].remove()\n",
    "\n",
    "            else:\n",
    "                print('sth. wrong')\n",
    "\n",
    "            for jj,_rgi_reg in enumerate(rgi_regs): \n",
    "                ax = axs[jj]\n",
    "                list_index_l = []\n",
    "                err_l = []\n",
    "                for m in path_d.keys():\n",
    "                    list_index_l.append(list(pd_cmip_temp_ch_2100.index + '_' + m))\n",
    "                pd_proj_2100_2300 = pd.DataFrame(index=np.concatenate(list_index_l))\n",
    "\n",
    "                if _rgi_reg !='global':\n",
    "                    rgi_reg = f'RGI{_rgi_reg}'\n",
    "                else:\n",
    "                    rgi_reg = _rgi_reg \n",
    "\n",
    "                ####\n",
    "                pd_fit_stats_reg = pd_fit_stats.loc[pd_fit_stats.region == rgi_reg]\n",
    "                pd_fit_stats_reg_lowess = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'lowess_predi']\n",
    "                pd_fit_stats_reg_exp = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'exp']\n",
    "                sel_it = pd_fit_stats_reg_lowess # we don't have it anymore ... .loc[pd_fit_stats_reg_lowess.it ==it]\n",
    "\n",
    "                ######\n",
    "                fig_fit, axs_fit = plt.subplots(2,1, figsize=(15,20)) \n",
    "\n",
    "                for end_year,ls in zip(['2300'],['-']):\n",
    "                    #for temp_lev in colors_temp.keys():\n",
    "\n",
    "                    for cmip in ['CMIP5','CMIP6']:\n",
    "                        pd_cmip = pd_cmip_temp_ch_2100_gcm_until_2300.loc[pd_cmip_temp_ch_2100_gcm_until_2300.ssp != 'ssp534-over']\n",
    "                        pd_cmip = pd_cmip.loc[pd_cmip.cmip == cmip]\n",
    "                        # check that there are temperature values in 2300 available ... \n",
    "                        pd_cmip = pd_cmip.loc[pd_cmip['global_temp_ch_2271-2300_preindustrial'].dropna().index]\n",
    "                        for gcm_scenario in pd_cmip.index:\n",
    "                            df_rel_l = []\n",
    "                            for m in path_d.keys():\n",
    "                                try:\n",
    "                                    gcm, scenario = gcm_scenario.split('_')\n",
    "                                    path = path_d[m]\n",
    "                                    gcm_scenario_m = gcm_scenario + '_' + m\n",
    "                                    n = len(gcm_scenario)\n",
    "                                    if _rgi_reg == '13-14-15':\n",
    "                                        df = 0\n",
    "                                        for sub_rgi_reg in ['RGI13','RGI14','RGI15']:\n",
    "                                            idir = path + f'volume/{cmip}/{end_year}/{sub_rgi_reg}/'\n",
    "                                            _df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                            _df.columns = [g.upper() for g in _df.columns]\n",
    "                                            _df = _df[gcm.upper()]\n",
    "                                            df += _df\n",
    "                                    else:\n",
    "                                        idir = path  + f'volume/{cmip}/{end_year}/{rgi_reg}/'\n",
    "                                        df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                        df.columns = [g.upper() for g in df.columns]\n",
    "                                        df = df[gcm.upper()]\n",
    "                                    df = df.loc[2000:]\n",
    "                                    # first do the rolling than the relative estimates .. \n",
    "                                    df_roll = df.rolling(20, center=True, min_periods=1).mean().dropna()\n",
    "                                    df_rel_roll = 100*df_roll/df_roll.loc[2020]                                    \n",
    "\n",
    "                                    temp_ch = pd_cmip.loc[gcm_scenario, 'global_temp_ch_2071-2100_preindustrial']\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, 'global_temp_ch_2071-2100_preindustrial'] = temp_ch\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2100_{rgi_reg}'] = df_rel_roll.loc[2100]\n",
    "\n",
    "                                    color= cmap(scaler.transform(temp_ch.reshape(-1,1))).flatten()\n",
    "                                    if not add_median:\n",
    "                                        alpha = 0.2\n",
    "                                        ax.plot(df_rel_roll.loc[2020:2300].index, df_rel_roll.loc[2020:2300], lw=2, color=color,ls=ls, alpha = alpha) \n",
    "                                        ax.plot(df_rel_roll.loc[2000:2020].index, df_rel_roll.loc[2000:2020], lw=3, color='black', ls=ls, alpha = alpha)\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2300_{rgi_reg}'] = df_rel_roll.loc[2300]\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, 'model'] = m\n",
    "\n",
    "                                    df_rel_l.append(df_rel_roll.loc[2000:2300])\n",
    "                                except:\n",
    "                                    pass\n",
    "                            if add_median:\n",
    "                                if len(df_rel_l) > 0:\n",
    "                                    df_rel_med = pd.DataFrame(df_rel_l).T.median(axis=1)\n",
    "                                    ax.plot(df_rel_med.loc[2020:2300].index, df_rel_med.loc[2020:2300], lw=2, color=color,ls=ls, alpha = 0.4) \n",
    "                                    ax.plot(df_rel_med.loc[2000:2020].index, df_rel_med.loc[2000:2020], lw=2, color='black', ls=ls, alpha = 0.2)\n",
    "\n",
    "                ax.set_ylim([0,125])\n",
    "                if jj in [0,2,4]:\n",
    "                    ax.set_ylabel('% of Remaining Ice',# (relative to 2020)',\n",
    "                                  fontweight = 'bold', fontsize=22)\n",
    "                else:\n",
    "                    ax.set_ylabel('')\n",
    "                ax.set_xlabel('Year',  fontweight = 'bold', fontsize=22)\n",
    "                handles, labels = ax.get_legend_handles_labels()\n",
    "                for h in handles:\n",
    "                    h.set_linewidth(3)\n",
    "                ax.legend(handles, labels,\n",
    "                              handlelength=1,\n",
    "                              title='Temperature change between 1850-1900 & 2071-2100:', ncol=5)\n",
    "                ax.legend().remove()\n",
    "\n",
    "                ax.set_title(d_reg_num_name_icci[_rgi_reg], fontweight='bold', fontsize=24)\n",
    "\n",
    "                pd_proj_2100_2300 = pd_proj_2100_2300.dropna()\n",
    "                assert len(pd_proj_2100_2300) == 60\n",
    "                for j,sim_year in enumerate([2100,2300]):\n",
    "                    # choose a \"frac\" parameter and select the right fitting values \n",
    "                    frac, _sel = select_frac_w_experts(sel_it, sim_year=sim_year)\n",
    "                    frac_dict[f'{rgi_reg}_{sim_year}'] = frac\n",
    "\n",
    "                    ax_fit = axs_fit[j]\n",
    "                    ax_fit.set_title(f'{d_reg_num_name_icci[_rgi_reg]}:  year {sim_year}', fontweight='bold')\n",
    "\n",
    "                    # get the raw data ...\n",
    "                    x = pd_proj_2100_2300['global_temp_ch_2071-2100_preindustrial']\n",
    "                    y = pd_proj_2100_2300[f'rel_ice_{sim_year}_{rgi_reg}']\n",
    "\n",
    "                    t_min, t_max = x.min(), x.max()\n",
    "                    if sim_year == 2100:\n",
    "                        delta_x = 0.002\n",
    "                    else:\n",
    "                        delta_x = 0.001\n",
    "\n",
    "                    df_quantiles = _sel[['q50','q05','q25','q75','q95']]\n",
    "                    # we just have to clip negative values .... there are still some although we tried to omit them by choosing the best frac ... \n",
    "                    df_quantiles = df_quantiles.where(df_quantiles>=0, 0)\n",
    "                    _sel_ice_eval = df_quantiles['q50'].values\n",
    "                    _sel_ice_eval_bottom = df_quantiles['q05'].values\n",
    "                    _sel_ice_eval_bottom_iqr = df_quantiles['q25'].values\n",
    "                    _sel_ice_eval_top_iqr = df_quantiles['q75'].values\n",
    "                    _sel_ice_eval_top = df_quantiles['q95'].values\n",
    "\n",
    "                    ax_fit.plot(pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].deltaTemp,\n",
    "                             pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].q50, lw=3, color='green', alpha = 0.5, label='exponential fit')\n",
    "                    ax_fit.plot(_sel.deltaTemp,\n",
    "                             _sel_ice_eval, lw=3, color='black', alpha = 1, label=f'lowess, frac={frac}')\n",
    "                    ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom, _sel_ice_eval_top, lw=3, alpha=0.2, color=f'black', label='90% PI')\n",
    "                    ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom_iqr, _sel_ice_eval_top_iqr, lw=3, alpha=0.4, color=f'black', label = '50% PI (IQR)')\n",
    "\n",
    "                    temp_eval_x = _sel.deltaTemp.values\n",
    "                    sns.scatterplot(ax=ax_fit,\n",
    "                                    data = pd_proj_2100_2300, #.loc[j], \n",
    "                                    x='global_temp_ch_2071-2100_preindustrial', y=f'rel_ice_{sim_year}_{rgi_reg}',\n",
    "                                                    style='model',\n",
    "                                    hue='global_temp_ch_2071-2100_preindustrial',\n",
    "                                    palette=cmap,      \n",
    "                                    sizes=[150,150,150], size='model') #, legend=)\n",
    "                    ax_fit.legend(fontsize=16, ncols=2)\n",
    "\n",
    "                    ax_fit.set_xlabel('Temperature change between 1850-1900 & 2071-2100')\n",
    "                    ax_fit.set_ylabel(f'% of Remaining Glacier Ice\\nin {sim_year} (relative to 2020)')\n",
    "\n",
    "                    for j,temp_lev in enumerate(list(colors_temp_values.keys())):\n",
    "\n",
    "                        # for the annotation in the main plot ... \n",
    "                        if temp_lev != '>4.0°C':\n",
    "\n",
    "                            temp_lev_value = colors_temp_values[temp_lev]\n",
    "                            color= cmap(scaler.transform(np.array([temp_lev_value]).reshape(-1,1))).flatten()\n",
    "                            idx_nearest = np.abs(temp_eval_x - temp_lev_value).argmin()\n",
    "                            if j >0:\n",
    "                                # check the previous fit, if they overlap, need to shift the text\n",
    "                                previous_fit = fit.copy()\n",
    "                            else:\n",
    "                                previous_fit = 10\n",
    "\n",
    "\n",
    "                            if previous_fit <4:\n",
    "                                shift = previous_fit-3.5\n",
    "                            else:\n",
    "                                shift = 0\n",
    "\n",
    "                            fit = _sel_ice_eval[idx_nearest]\n",
    "\n",
    "\n",
    "\n",
    "                            if sim_year == 2300:\n",
    "                                loc_2300_stats = 2300 +4 \n",
    "                            else:\n",
    "                                loc_2300_stats = 2100\n",
    "                            #if temp_lev_value>1.5:\n",
    "                            #    ax.vlines(loc_2300_stats,y_fitted[0], y_fitted[1], color=colors_temp[temp_lev], lw=15, alpha = 0.4)\n",
    "                            ax.hlines(fit,loc_2300_stats-3, loc_2300_stats+3, lw=8, color=color, zorder=10)\n",
    "\n",
    "                            if sim_year == 2300:\n",
    "                                bbox=dict(facecolor='white', edgecolor='white', boxstyle='round', alpha = 0.8)\n",
    "                                ax.annotate(temp_lev[2:], xy=(2300, fit),\n",
    "                                            xytext=(loc_2300_stats+delta_text, fit+shift),\n",
    "                                                        verticalalignment=\"center\",\n",
    "                                                        color=color, alpha=1, fontweight = 'bold', fontsize=18)\n",
    "                                _sel_temp_lev_qs = np.array([ _sel_ice_eval_bottom[idx_nearest],\n",
    "                                                             _sel_ice_eval_bottom_iqr[idx_nearest], _sel_ice_eval[idx_nearest],\n",
    "                                                             _sel_ice_eval_top_iqr[idx_nearest], _sel_ice_eval_top[idx_nearest]])\n",
    "                                if uncertainty == 'boxplot':\n",
    "                                    boxplot_manual(qs= _sel_temp_lev_qs,\n",
    "                                                   p = loc_2300_stats+5.5,\n",
    "                                                   ax = ax, j=j*4.1, temp_val = temp_lev_value,\n",
    "                                                   )\n",
    "                    np.testing.assert_allclose(temp_eval_x[::2], np.arange(1.4,6.3,0.1))\n",
    "                    colors_fit = cmap(scaler.transform(temp_eval_x[::2].reshape(-1,1))) # only every decile ... \n",
    "                    lw = 2\n",
    "                    for c,f in zip(colors_fit, _sel_ice_eval[::2]):\n",
    "                        ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                                   color=c, zorder=7)\n",
    "\n",
    "                ax.set_xlim([1997, xlim1])\n",
    "                if rgi_reg == 'RGI08':\n",
    "                    ax_fit.set_ylim([-0.5, 120])\n",
    "\n",
    "                ax.set_xticks([2020,2100,2200,2300],[2020,2100,2200,2300])\n",
    "                ax.spines['right'].set_visible(False)\n",
    "                ax.spines['top'].set_visible(False)\n",
    "\n",
    "                fig_fit.tight_layout()\n",
    "                if uncertainty == 'boxplot':\n",
    "                        fig_fit.savefig(f'figures/appendix_fit/lowess_predi_exp_fit_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3.png')\n",
    "\n",
    "            ax_f.text(2290,119, 'colors describe global\\ntemperature change in 2100',\n",
    "                                    #transform=ax_f.transAxes,\n",
    "                                    verticalalignment=\"top\", bbox=props, horizontalalignment='right',\n",
    "                                    color='grey', alpha=1, fontweight = 'bold', fontsize=18)\n",
    "            fig.tight_layout()\n",
    "            # for every of the four large regions ... \n",
    "            if reg == 'Glaciers globally':\n",
    "                reg = 'Glaciers-globally'\n",
    "            if uncertainty== 'boxplot':\n",
    "\n",
    "                if add_median:\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/png/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_v3_{uncertainty}.png')\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/svg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_v3_{uncertainty}.svg')\n",
    "                else:\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/png/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_v3_{uncertainty}.png')\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/svg/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_v3_{uncertainty}.svg')\n",
    "\n",
    "            else:\n",
    "                if add_median:\n",
    "                    fig.savefig(f'figures/main_figures_without_uncertainties/median_glac_models/png/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_v3_{uncertainty}.png')\n",
    "                    fig.savefig(f'figures/main_figures_without_uncertainties/median_glac_models/svg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_v3_{uncertainty}.svg')\n",
    "                else:\n",
    "                    fig.savefig(f'figures/main_figures_without_uncertainties/all_glac_models/png/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_v3_{uncertainty}.png')\n",
    "                    fig.savefig(f'figures/main_figures_without_uncertainties/all_glac_models/svg/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_v3_{uncertainty}.svg')\n",
    "\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6751ce23",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "39c73bff",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "3cd2578f",
   "metadata": {},
   "source": [
    "#### **Final versions of the figures for the ICCI State of the Cryosphere report 2023**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "0a634145",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/html": [
       "<div style=\"vertical-align: middle;\"><strong>bins_100</strong> </div><div class=\"cmap\"><img alt=\"bins_100 colormap\" title=\"bins_100\" style=\"border: 1px solid #555;\" src=\"data:image/png;base64,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\"></div><div style=\"vertical-align: middle; max-width: 514px; display: flex; justify-content: space-between;\"><div style=\"float: left;\"><div title=\"#70b8ffff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #70b8ffff;\"></div> under</div><div style=\"margin: 0 auto; display: inline-block;\">bad <div title=\"#00000000\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #00000000;\"></div></div><div style=\"float: right;\">over <div title=\"#662b00ff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #662b00ff;\"></div></div>"
      ],
      "text/plain": [
       "<matplotlib.colors.LinearSegmentedColormap at 0x7ff96219f950>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "colors_icci_l1 = ['#8AE6E6', '#FFAA00', '#F27900', '#D93D15', '#B3001E', '#800040', '#550066', '#2B084D']\n",
    "colors_icci_l2 = ['#99DDFF', '#FABB00', '#FA8800', '#E64D17', '#BF262D', '#8C0053', '#581466', '#310959']\n",
    "colors_icci_l2_sh = colors_icci_l2[:-2]\n",
    "\n",
    "colors_icci_l3 = ['#70B8FF', '#FABB00', '#F25100', '#D42300', '#B3001E', '#800040', '#550066', '#2B084D']\n",
    "cmap_icci_3 = LinearSegmentedColormap.from_list('bins_100', colors_icci_l3, N=10000)\n",
    "\n",
    "colors_icci_below3deg = ['#70B8FF', '#FABB00', '#FF8C00', '#F25100', '#BD2600', '#662B00'] #, '#662B00', '#662B00']\n",
    "cmap_icci_below3deg = LinearSegmentedColormap.from_list('bins_100', colors_icci_below3deg, N=10000)\n",
    "\n",
    "cmap_icci_below3deg\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "4f6b865c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/html": [
       "<div style=\"vertical-align: middle;\"><strong>bins_100</strong> </div><div class=\"cmap\"><img alt=\"bins_100 colormap\" title=\"bins_100\" style=\"border: 1px solid #555;\" src=\"data:image/png;base64,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\"></div><div style=\"vertical-align: middle; max-width: 514px; display: flex; justify-content: space-between;\"><div style=\"float: left;\"><div title=\"#8ae6e6ff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #8ae6e6ff;\"></div> under</div><div style=\"margin: 0 auto; display: inline-block;\">bad <div title=\"#00000000\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #00000000;\"></div></div><div style=\"float: right;\">over <div title=\"#2b084dff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #2b084dff;\"></div></div>"
      ],
      "text/plain": [
       "<matplotlib.colors.LinearSegmentedColormap at 0x7ff9bafea750>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmap_icci_1 = LinearSegmentedColormap.from_list('bins_100', colors_icci_l1, N=10000)\n",
    "cmap_icci_1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "b8de6997",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/html": [
       "<div style=\"vertical-align: middle;\"><strong>bins_100</strong> </div><div class=\"cmap\"><img alt=\"bins_100 colormap\" title=\"bins_100\" style=\"border: 1px solid #555;\" src=\"data:image/png;base64,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\"></div><div style=\"vertical-align: middle; max-width: 514px; display: flex; justify-content: space-between;\"><div style=\"float: left;\"><div title=\"#99ddffff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #99ddffff;\"></div> under</div><div style=\"margin: 0 auto; display: inline-block;\">bad <div title=\"#00000000\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #00000000;\"></div></div><div style=\"float: right;\">over <div title=\"#8c0053ff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #8c0053ff;\"></div></div>"
      ],
      "text/plain": [
       "<matplotlib.colors.LinearSegmentedColormap at 0x7ff9ba72d3d0>"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cmap_icci_2 = LinearSegmentedColormap.from_list('bins_100', colors_icci_l2, N=1000)\n",
    "cmap_icci_2\n",
    "\n",
    "cmap_icci_2_sh = LinearSegmentedColormap.from_list('bins_100', colors_icci_l2_sh, N=1000)\n",
    "cmap_icci_2_sh\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "2c606d02",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgAAAABACAYAAABsv8+/AAAAF3RFWHRUaXRsZQBiaW5zXzEwMCBjb2xvcm1hcLv0aLkAAAAddEVYdERlc2NyaXB0aW9uAGJpbnNfMTAwIGNvbG9ybWFwkoqPkAAAADB0RVh0QXV0aG9yAE1hdHBsb3RsaWIgdjMuNy4xLCBodHRwczovL21hdHBsb3RsaWIub3JnKq/4TQAAADJ0RVh0U29mdHdhcmUATWF0cGxvdGxpYiB2My43LjEsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcECWdqAAABrklEQVR4nO3WUU6DQBSG0TvQpboll+kGOgI+lBqCmpakapr/nIQA05uZhqevvby+La2mOg29xnauceh1ar3G1j/XLve+/raZWde/XRvONbZeQ5s2e/Wv51z33Mz8eM6653VtqLlqqap5vW49H509Mn/k/KWe11JVU1W9r/d7r3vm5wfutZ9/4m8+767pwPut2bnaw/bavy/VfuV7AI8x/PcfAAD+ngAAgEACAAACCQAACCQAACCQAACAQAIAAAIJAAAIJAAAIJAAAIBAAgAAAgkAAAgkAAAgkAAAgEACAAACCQAACCQAACCQAACAQAIAAAIJAAAIJAAAIJAAAIBAAgAAAgkAAAgkAAAgkAAAgEACAAACCQAACCQAACCQAACAQAIAAAIJAAAIJAAAIJAAAIBAAgAAAgkAAAgkAAAgkAAAgEACAAACCQAACCQAACCQAACAQAIAAAIJAAAIJAAAIJAAAIBAAgAAAgkAAAgkAAAgkAAAgEACAAACCQAACCQAACCQAACAQAIAAAIJAAAIJAAAIJAAAIBAAgAAAgkAAAgkAAAgkAAAgEACAAACfQAab70ZZWAp5gAAAABJRU5ErkJggg==",
      "text/html": [
       "<div style=\"vertical-align: middle;\"><strong>bins_100</strong> </div><div class=\"cmap\"><img alt=\"bins_100 colormap\" title=\"bins_100\" style=\"border: 1px solid #555;\" src=\"data:image/png;base64,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\"></div><div style=\"vertical-align: middle; max-width: 514px; display: flex; justify-content: space-between;\"><div style=\"float: left;\"><div title=\"#6495edff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #6495edff;\"></div> under</div><div style=\"margin: 0 auto; display: inline-block;\">bad <div title=\"#00000000\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #00000000;\"></div></div><div style=\"float: right;\">over <div title=\"#980002ff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #980002ff;\"></div></div>"
      ],
      "text/plain": [
       "<matplotlib.colors.LinearSegmentedColormap at 0x7ff9a787b7d0>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "colors_temp5_l = ['cornflowerblue', 'gold', 'darkorange','darkorange', 'red','red', '#980002', '#980002','#980002', '#980002', '#980002', '#980002',]\n",
    "cmap_blue_red = LinearSegmentedColormap.from_list('bins_100',colors_temp5_l, N=1000)\n",
    "cmap_blue_red"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "bb901904",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/html": [
       "<div style=\"vertical-align: middle;\"><strong>bins_100</strong> </div><div class=\"cmap\"><img alt=\"bins_100 colormap\" title=\"bins_100\" style=\"border: 1px solid #555;\" src=\"data:image/png;base64,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\"></div><div style=\"vertical-align: middle; max-width: 514px; display: flex; justify-content: space-between;\"><div style=\"float: left;\"><div title=\"#6495edff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #6495edff;\"></div> under</div><div style=\"margin: 0 auto; display: inline-block;\">bad <div title=\"#00000000\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #00000000;\"></div></div><div style=\"float: right;\">over <div title=\"#980002ff\" style=\"display: inline-block; width: 1em; height: 1em; margin: 0; vertical-align: middle; border: 1px solid #555; background-color: #980002ff;\"></div></div>"
      ],
      "text/plain": [
       "<matplotlib.colors.LinearSegmentedColormap at 0x7ffa0258e450>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from matplotlib.colors import LinearSegmentedColormap\n",
    "colors_temp6_l = ['cornflowerblue', 'gold', 'darkorange', 'red','#980002']\n",
    "cmap_blue_red_bel3deg = LinearSegmentedColormap.from_list('bins_100',colors_temp6_l, N=1000)\n",
    "cmap_blue_red_bel3deg"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "80a63cda",
   "metadata": {},
   "source": [
    "**another scaler only going until 3 degree:**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "72a14bd9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.427504187860568 2.907176938010541\n"
     ]
    }
   ],
   "source": [
    "pd_cmip_sel_below3deg = pd_cmip_temp_ch_2100_gcm_until_2300.loc[pd_cmip_temp_ch_2100_gcm_until_2300['global_temp_ch_2071-2100_preindustrial']<=3]\n",
    "data_temp_bel3deg = pd_cmip_sel_below3deg['global_temp_ch_2071-2100_preindustrial'].values.reshape(1,-1).T\n",
    "scaler_below3deg = MinMaxScaler()\n",
    "scaler_below3deg.fit(data_temp_bel3deg)\n",
    "t_min_bel3deg, t_max_bel3deg = pd_cmip_sel_below3deg['global_temp_ch_2071-2100_preindustrial'].min(), pd_cmip_sel_below3deg['global_temp_ch_2071-2100_preindustrial'].max()\n",
    "print(t_min_bel3deg, t_max_bel3deg)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b8eb756",
   "metadata": {},
   "source": [
    "**GCMs going below / equal 3 degree**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "3283d17e",
   "metadata": {},
   "outputs": [],
   "source": [
    "temp_eval_x = np.arange(1.4,6.3,0.001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "acef513c",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_90224/677254350.py:22: UserWarning: Tight layout not applied. The bottom and top margins cannot be made large enough to accommodate all axes decorations.\n",
      "  plt.tight_layout()\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x100 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(4,1))\n",
    "cmap = cmap_icci_below3deg\n",
    "_scaler = scaler_below3deg\n",
    "temp_eval_x_sel = temp_eval_x[temp_eval_x<=3]\n",
    "colors_fit = cmap(_scaler.transform(temp_eval_x_sel.reshape(-1,1))) # only every decile ... \n",
    "for f,c in zip(temp_eval_x_sel, colors_fit):\n",
    "    plt.vlines(f,1,1.2, lw=3, alpha = 0.5,\n",
    "               color=c, zorder=7)\n",
    "    \n",
    "for t in pd_cmip_sel_below3deg['global_temp_ch_2071-2100_preindustrial'].values:\n",
    "    c = cmap(_scaler.transform(t.reshape(-1,1))) # only every decile ... \n",
    "\n",
    "    plt.vlines(t,0.95,1, lw=1, color=c)\n",
    "plt.yticks([])\n",
    "ax = plt.gca()\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['left'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(False)\n",
    "plt.xticks([1.5,2,2.5,3], [1.5,2,2.5,3], fontsize=15, color='grey')\n",
    "plt.ylim([0.95,1.2])\n",
    "plt.tight_layout()\n",
    "plt.savefig('figures/colorbar.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "d42ac10e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_90224/2219712573.py:22: UserWarning: Tight layout not applied. The bottom and top margins cannot be made large enough to accommodate all axes decorations.\n",
      "  plt.tight_layout()\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x100 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(5,1))\n",
    "cmap = cmap_icci_below3deg\n",
    "_scaler = scaler_below3deg\n",
    "temp_eval_x_sel = temp_eval_x[temp_eval_x<=3]\n",
    "colors_fit = cmap(_scaler.transform(temp_eval_x_sel.reshape(-1,1))) # only every decile ... \n",
    "for f,c in zip(temp_eval_x_sel, colors_fit):\n",
    "    plt.vlines(f,1,1.2, lw=3, alpha = 0.5,\n",
    "               color=c, zorder=7)\n",
    "    \n",
    "for t in pd_cmip_sel_below3deg['global_temp_ch_2071-2100_preindustrial'].values:\n",
    "    c = cmap(_scaler.transform(t.reshape(-1,1))) # only every decile ... \n",
    "\n",
    "    plt.vlines(t,0.95,1, lw=1, color=c)\n",
    "plt.yticks([])\n",
    "ax = plt.gca()\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['left'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(False)\n",
    "plt.xticks(fontsize=15, color='grey')\n",
    "plt.ylim([0.95,1.2])\n",
    "plt.tight_layout()\n",
    "\n",
    "plt.savefig('figures/colorbar2.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "504ca1d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "# this file has been merged in lowess_percentile_interval_fit_create_fit_data_and_merge_lowess_final_fit_files.ipynb\n",
    "# the actual fits were computed by lowess_percentile_interval_fit_per_region.py\n",
    "pd_fit_stats = pd.read_csv('lowess_fit_stats_oct29_predi_all_it1.csv', index_col=0)\n",
    "def boxplot_manual(qs= [0.1,0.5,1,2,3],\n",
    "                   p = 2300, ax = None, j=0, temp_val = 1.5, _scaler=scaler):\n",
    "    q05, q25, q50, q75, q95 = qs\n",
    "    if temp_val == None:\n",
    "        c = 'grey'\n",
    "    else:\n",
    "        c = cmap(_scaler.transform(np.array([temp_val]).reshape(-1,1))).flatten()\n",
    "    item = {}\n",
    "    #item[\"label\"] = 'box' # not required\n",
    "    item[\"med\"] = q50\n",
    "    item[\"q1\"] = q25\n",
    "    item[\"q3\"] = q75\n",
    "    #item[\"cilo\"] = 5.3 # not required\n",
    "    #item[\"cihi\"] = 5.7 # not required\n",
    "    item[\"whislo\"] = q05\n",
    "    item[\"whishi\"] = q95\n",
    "    stats = [item]\n",
    "    boxprops = dict(linestyle='-', linewidth=2, color=c)\n",
    "    whisprops = dict(color=c, linewidth=1)\n",
    "    flierprops = dict(marker='o', markerfacecolor='green', markersize=0,\n",
    "                      markeredgecolor='none')\n",
    "    medianprops = dict(linestyle='-', linewidth=4, color=c)\n",
    "    ax.bxp(stats, positions = [p + j], showfliers = False,\n",
    "            flierprops= flierprops, medianprops=medianprops, boxprops=boxprops,\n",
    "             widths=4.5, #patch_artist=True,\n",
    "            whiskerprops=whisprops, capprops = dict(color=c, linewidth=1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "id": "6459e81a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI16 0.41 0 0.5940125822019444 -0.009331074123102\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI08 0.96 0 1.0382209879829225 0.0437381748806752\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI11 0.51 0 0.4329462619194236 -0.0115070402162595\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI02 0.46 0 0.0863687406451393 -0.0073848595658214\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI18 0.86 0 2.421712130179934 0.0639439491791801\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI06 0.99 0 21.31067898391548 0.075307676741005\n",
      "RGI06 problem\n",
      "RGI06 0.8 0 -1.6090488460476096 -0.2584169616942668\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13-14-15 0.99 0 8.45265502846896 0.0300477867204858\n",
      "RGI13-14-15 0.57 0 -0.0536211913880723 -0.3423459518639633\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI01 0.99 0 24.468128333237928 0.0562669460517852\n",
      "RGI01 0.57 0 0.4378871108982666 -0.0461755604138239\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI04 0.97 0 17.45129269512268 -0.2584482711844496\n",
      "RGI04 0.72 -0.5 -1.453974934319801 -0.1119918149230363\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI17 0.99 0 16.27953236489871 0.3190121418276295\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "global 0.99 0 40.35323776152246 0.1281386443393088\n",
      "global 0.45 0 0.4881256712591804 -0.18261132766947\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI16 0.41 0 0.5940125822019444 -0.009331074123102\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI08 0.96 0 1.0382209879829225 0.0437381748806752\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI11 0.51 0 0.4329462619194236 -0.0115070402162595\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI02 0.46 0 0.0863687406451393 -0.0073848595658214\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI18 0.86 0 2.421712130179934 0.0639439491791801\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI06 0.99 0 21.31067898391548 0.075307676741005\n",
      "RGI06 problem\n",
      "RGI06 0.8 0 -1.6090488460476096 -0.2584169616942668\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13-14-15 0.99 0 8.45265502846896 0.0300477867204858\n",
      "RGI13-14-15 0.57 0 -0.0536211913880723 -0.3423459518639633\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI01 0.99 0 24.468128333237928 0.0562669460517852\n",
      "RGI01 0.57 0 0.4378871108982666 -0.0461755604138239\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI04 0.97 0 17.45129269512268 -0.2584482711844496\n",
      "RGI04 0.72 -0.5 -1.453974934319801 -0.1119918149230363\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI17 0.99 0 16.27953236489871 0.3190121418276295\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "global 0.99 0 40.35323776152246 0.1281386443393088\n",
      "global 0.45 0 0.4881256712591804 -0.18261132766947\n"
     ]
    }
   ],
   "source": [
    "v = 3 ## COLOR SCALE version ... \n",
    "add_median = True\n",
    "\n",
    "for only_below3deg in [True,False]:\n",
    "\n",
    "    if only_below3deg:\n",
    "        vv = f'v{v}_below3deg'\n",
    "        _scaler = scaler_below3deg\n",
    "        if v == 'm':\n",
    "            cmap = cmap_blue_red_bel3deg\n",
    "        elif v == 1:\n",
    "            cmap = cmap_icci_1\n",
    "        elif v==2:\n",
    "            cmap = cmap_icci_2_sh\n",
    "        elif v==3:\n",
    "            cmap = cmap_icci_below3deg\n",
    "        caption_text = 'colors describe global temperature\\nchange in 2100 from +1.4°C to +3.0°C'\n",
    "\n",
    "    else:\n",
    "        vv = f'v{v}_all'\n",
    "        _scaler = scaler\n",
    "        #cmap = cmap_icci_1 #cmap_icci_2#cmap_blue_red\n",
    "        if v == 'm':\n",
    "            cmap = cmap_blue_red\n",
    "        elif v == 1:\n",
    "            cmap = cmap_icci_1\n",
    "        elif v==2:\n",
    "            cmap = cmap_icci_2\n",
    "        elif v==3:\n",
    "            cmap = cmap_icci_3\n",
    "\n",
    "        caption_text = 'colors describe global temperature\\nchange in 2100 from +1.4°C to +6.3°C'\n",
    "    color_continuous = True\n",
    "    decile_ticks = True\n",
    "    plt.rc('font', size=26)\n",
    "    it = 1\n",
    "    frac_dict = {}\n",
    "    for uncertainty in [ 'boxplot']: #, 'no_uncertainties']: # 'no_uncertainties',\n",
    "        if uncertainty == 'boxplot':\n",
    "            num_fits = 1000\n",
    "            delta_text = 23\n",
    "            xlim1 = 2327\n",
    "        else:\n",
    "            num_fits = 1000 \n",
    "            delta_text = 4\n",
    "            xlim1 = 2309\n",
    "\n",
    "        for reg in list(d_reg_num_name_icci_reg.keys()):\n",
    "            rgi_regs = d_reg_num_name_icci_reg[reg]\n",
    "\n",
    "            len_sub_p = len(rgi_regs)\n",
    "            props = dict(boxstyle='round', facecolor='white', alpha=0.5)\n",
    "            if (reg == 'Glaciers globally') or (reg == 'low-latitudes'):\n",
    "                fig, axs = plt.subplots(1,1, figsize=(12.5,25/3)) \n",
    "                axs = [axs]#.flatten()\n",
    "                #for a in [1,2,3,4,5]:\n",
    "                #    axs[a].remove()\n",
    "                # figure to show caption \n",
    "                ax_f = axs[0] \n",
    "\n",
    "            elif reg == 'mid-latitudes':\n",
    "                fig, axs = plt.subplots(2,2, figsize=(25,25*2/3))\n",
    "                axs = axs.flatten()\n",
    "                #axs[4].remove()\n",
    "                #axs[5].remove()\n",
    "                ax_f = axs[2]\n",
    "            elif reg == 'polar-hma':\n",
    "                fig, axs = plt.subplots(3,2, figsize=(25,25))\n",
    "                axs = axs.flatten()\n",
    "                ax_f = axs[4]\n",
    "                #ax_f.spines['left'].set_visible(False)\n",
    "                #ax_f.spines['bottom'].set_visible(False)\n",
    "                axs[5].remove()\n",
    "\n",
    "            else:\n",
    "                print('sth. wrong')\n",
    "\n",
    "            for jj,_rgi_reg in enumerate(rgi_regs): \n",
    "                ax = axs[jj]\n",
    "                list_index_l = []\n",
    "                err_l = []\n",
    "                for m in path_d.keys():\n",
    "                    list_index_l.append(list(pd_cmip_temp_ch_2100.index + '_' + m))\n",
    "                pd_proj_2100_2300 = pd.DataFrame(index=np.concatenate(list_index_l))\n",
    "\n",
    "                if _rgi_reg !='global':\n",
    "                    rgi_reg = f'RGI{_rgi_reg}'\n",
    "                else:\n",
    "                    rgi_reg = _rgi_reg \n",
    "\n",
    "                ####\n",
    "                pd_fit_stats_reg = pd_fit_stats.loc[pd_fit_stats.region == rgi_reg]\n",
    "                pd_fit_stats_reg_lowess = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'lowess_predi']\n",
    "                pd_fit_stats_reg_exp = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'exp']\n",
    "                sel_it = pd_fit_stats_reg_lowess # we don't have it anymore ... .loc[pd_fit_stats_reg_lowess.it ==it]\n",
    "\n",
    "                ######\n",
    "                fig_fit, axs_fit = plt.subplots(2,1, figsize=(15,20)) \n",
    "\n",
    "                for end_year,ls in zip(['2300'],['-']):\n",
    "                    #for temp_lev in colors_temp.keys():\n",
    "\n",
    "                    for cmip in ['CMIP5','CMIP6']:\n",
    "                        pd_cmip = pd_cmip_temp_ch_2100_gcm_until_2300.loc[pd_cmip_temp_ch_2100_gcm_until_2300.ssp != 'ssp534-over']\n",
    "                        pd_cmip = pd_cmip.loc[pd_cmip.cmip == cmip]\n",
    "                        # check that there are temperature values in 2300 available ... \n",
    "                        pd_cmip = pd_cmip.loc[pd_cmip['global_temp_ch_2271-2300_preindustrial'].dropna().index]\n",
    "                        for gcm_scenario in pd_cmip.index:\n",
    "                            df_rel_l = []\n",
    "                            for m in path_d.keys():\n",
    "                                try:\n",
    "                                    gcm, scenario = gcm_scenario.split('_')\n",
    "                                    path = path_d[m]\n",
    "                                    gcm_scenario_m = gcm_scenario + '_' + m\n",
    "                                    n = len(gcm_scenario)\n",
    "                                    if _rgi_reg == '13-14-15':\n",
    "                                        df = 0\n",
    "                                        for sub_rgi_reg in ['RGI13','RGI14','RGI15']:\n",
    "                                            idir = path + f'volume/{cmip}/{end_year}/{sub_rgi_reg}/'\n",
    "                                            _df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                            _df.columns = [g.upper() for g in _df.columns]\n",
    "                                            _df = _df[gcm.upper()]\n",
    "                                            df += _df\n",
    "                                    else:\n",
    "                                        idir = path  + f'volume/{cmip}/{end_year}/{rgi_reg}/'\n",
    "                                        df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                        df.columns = [g.upper() for g in df.columns]\n",
    "                                        df = df[gcm.upper()]\n",
    "                                    df = df.loc[2000:]\n",
    "                                    # first do the rolling than the relative estimates .. \n",
    "                                    df_roll = df.rolling(20, center=True, min_periods=1).mean().dropna()\n",
    "                                    df_rel_roll = 100*df_roll/df_roll.loc[2020]                                    \n",
    "\n",
    "                                    temp_ch = pd_cmip.loc[gcm_scenario, 'global_temp_ch_2071-2100_preindustrial']\n",
    "                                    # if temp_ch < 3:\n",
    "                                    #if only_below3deg:\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, 'global_temp_ch_2071-2100_preindustrial'] = temp_ch\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2100_{rgi_reg}'] = df_rel_roll.loc[2100]\n",
    "\n",
    "                                    color= cmap(_scaler.transform(temp_ch.reshape(-1,1))).flatten()\n",
    "                                    if not add_median:\n",
    "                                        alpha = 0.2\n",
    "                                        ax.plot(df_rel_roll.loc[2020:2300].index, df_rel_roll.loc[2020:2300], lw=2, color=color,ls=ls, alpha = alpha) \n",
    "                                        ax.plot(df_rel_roll.loc[2000:2020].index, df_rel_roll.loc[2000:2020], lw=3, color='black', ls=ls, alpha = alpha)\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2300_{rgi_reg}'] = df_rel_roll.loc[2300]\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, 'model'] = m\n",
    "\n",
    "                                    df_rel_l.append(df_rel_roll.loc[2000:2300])\n",
    "                                except:\n",
    "                                    pass\n",
    "                            if add_median:\n",
    "                                if len(df_rel_l) > 0:\n",
    "                                    if (not only_below3deg) or (temp_ch<=3):\n",
    "                                        df_rel_med = pd.DataFrame(df_rel_l).T.median(axis=1)\n",
    "                                        ax.plot(df_rel_med.loc[2020:2300].index, df_rel_med.loc[2020:2300], lw=2, color=color,ls=ls, alpha = 0.3) \n",
    "                                        ax.plot(df_rel_med.loc[2000:2020].index, df_rel_med.loc[2000:2020], lw=2, color='black', ls=ls, alpha = 0.3)\n",
    "\n",
    "                ax.set_ylim([0,125])\n",
    "                if jj in [0,2,4]:\n",
    "                    ax.set_ylabel('% of Remaining Ice',# (relative to 2020)',\n",
    "                                  fontweight = 'bold', fontsize=22)\n",
    "                else:\n",
    "                    ax.set_ylabel('')\n",
    "                ax.set_xlabel('Year',  fontweight = 'bold', fontsize=22)\n",
    "                handles, labels = ax.get_legend_handles_labels()\n",
    "                for h in handles:\n",
    "                    h.set_linewidth(3)\n",
    "                ax.legend(handles, labels,\n",
    "                              handlelength=1,\n",
    "                              title='Temperature change between 1850-1900 & 2071-2100:', ncol=5)\n",
    "                ax.legend().remove()\n",
    "\n",
    "                ax.set_title(d_reg_num_name_icci[_rgi_reg], fontweight='bold', fontsize=24)\n",
    "\n",
    "                pd_proj_2100_2300 = pd_proj_2100_2300.dropna()\n",
    "                assert len(pd_proj_2100_2300) == 60\n",
    "                for j,sim_year in enumerate([2100,2300]):\n",
    "                    # choose a \"frac\" parameter and select the right fitting values \n",
    "                    frac, _sel = select_frac_w_experts(sel_it, sim_year=sim_year)\n",
    "                    frac_dict[f'{rgi_reg}_{sim_year}'] = frac\n",
    "\n",
    "                    ax_fit = axs_fit[j]\n",
    "                    ax_fit.set_title(f'{d_reg_num_name_icci[_rgi_reg]}:  year {sim_year}', fontweight='bold')\n",
    "\n",
    "                    # get the raw data ...\n",
    "                    x = pd_proj_2100_2300['global_temp_ch_2071-2100_preindustrial']\n",
    "                    y = pd_proj_2100_2300[f'rel_ice_{sim_year}_{rgi_reg}']\n",
    "\n",
    "                    t_min, t_max = x.min(), x.max()\n",
    "                    if sim_year == 2100:\n",
    "                        delta_x = 0.002\n",
    "                    else:\n",
    "                        delta_x = 0.001\n",
    "\n",
    "                    df_quantiles = _sel[['q50','q05','q25','q75','q95']]\n",
    "                    # we just have to clip negative values .... there are still some although we tried to omit them by choosing the best frac ... \n",
    "                    df_quantiles = df_quantiles.where(df_quantiles>=0, 0)\n",
    "                    _sel_ice_eval = df_quantiles['q50'].values\n",
    "                    _sel_ice_eval_bottom = df_quantiles['q05'].values\n",
    "                    _sel_ice_eval_bottom_iqr = df_quantiles['q25'].values\n",
    "                    _sel_ice_eval_top_iqr = df_quantiles['q75'].values\n",
    "                    _sel_ice_eval_top = df_quantiles['q95'].values\n",
    "\n",
    "                    ax_fit.plot(pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].deltaTemp,\n",
    "                             pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].q50, lw=3, color='green', alpha = 0.5, label='exponential fit')\n",
    "                    ax_fit.plot(_sel.deltaTemp,\n",
    "                             _sel_ice_eval, lw=3, color='black', alpha = 1, label=f'lowess, frac={frac}')\n",
    "                    ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom, _sel_ice_eval_top, lw=3, alpha=0.2, color=f'black', label='90% PI')\n",
    "                    ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom_iqr, _sel_ice_eval_top_iqr, lw=3, alpha=0.4, color=f'black', label = '50% PI (IQR)')\n",
    "\n",
    "                    temp_eval_x = _sel.deltaTemp.values\n",
    "                    sns.scatterplot(ax=ax_fit,\n",
    "                                    data = pd_proj_2100_2300, #.loc[j], \n",
    "                                    x='global_temp_ch_2071-2100_preindustrial', y=f'rel_ice_{sim_year}_{rgi_reg}',\n",
    "                                                    style='model',\n",
    "                                    hue='global_temp_ch_2071-2100_preindustrial',\n",
    "                                    palette=cmap,      \n",
    "                                    sizes=[150,150,150], size='model') #, legend=)\n",
    "                    ax_fit.legend(fontsize=16, ncols=2)\n",
    "\n",
    "                    ax_fit.set_xlabel('Temperature change between 1850-1900 & 2071-2100')\n",
    "                    ax_fit.set_ylabel(f'% of Remaining Glacier Ice\\nin {sim_year} (relative to 2020)')\n",
    "\n",
    "                    for j,temp_lev in enumerate(list(colors_temp_values.keys())):\n",
    "\n",
    "                        # for the annotation in the main plot ... \n",
    "                        if (temp_lev != '>4.0°C') and (temp_lev != '<=+4.0°C'):\n",
    "\n",
    "                            temp_lev_value = colors_temp_values[temp_lev]\n",
    "                            color= cmap(_scaler.transform(np.array([temp_lev_value]).reshape(-1,1))).flatten()\n",
    "                            idx_nearest = np.abs(temp_eval_x - temp_lev_value).argmin()\n",
    "                            if j >0:\n",
    "                                # check the previous fit, if they overlap, need to shift the text\n",
    "                                previous_fit = fit.copy()\n",
    "                            else:\n",
    "                                previous_fit = 10\n",
    "\n",
    "\n",
    "                            if previous_fit <4:\n",
    "                                shift = previous_fit-3.5\n",
    "                            else:\n",
    "                                shift = 0\n",
    "\n",
    "                            fit = _sel_ice_eval[idx_nearest]\n",
    "\n",
    "\n",
    "\n",
    "                            if sim_year == 2300:\n",
    "                                loc_2300_stats = 2300 +4 \n",
    "                            else:\n",
    "                                loc_2300_stats = 2100\n",
    "                            #if temp_lev_value>1.5:\n",
    "                            #    ax.vlines(loc_2300_stats,y_fitted[0], y_fitted[1], color=colors_temp[temp_lev], lw=15, alpha = 0.4)\n",
    "                            ax.hlines(fit,loc_2300_stats-3, loc_2300_stats+3, lw=8, color=color, zorder=10)\n",
    "\n",
    "                            if sim_year == 2300:\n",
    "                                bbox=dict(facecolor='white', edgecolor='white', boxstyle='round', alpha = 0.8)\n",
    "                                ax.annotate(temp_lev[2:], xy=(2300, fit),\n",
    "                                            xytext=(loc_2300_stats+delta_text, fit+shift),\n",
    "                                                        verticalalignment=\"center\",\n",
    "                                                        color=color, alpha=1, fontweight = 'bold', fontsize=18)\n",
    "                                _sel_temp_lev_qs = np.array([ _sel_ice_eval_bottom[idx_nearest],\n",
    "                                                             _sel_ice_eval_bottom_iqr[idx_nearest], _sel_ice_eval[idx_nearest],\n",
    "                                                             _sel_ice_eval_top_iqr[idx_nearest], _sel_ice_eval_top[idx_nearest]])\n",
    "                                if uncertainty == 'boxplot':\n",
    "                                    boxplot_manual(qs= _sel_temp_lev_qs,\n",
    "                                                   p = loc_2300_stats+7,\n",
    "                                                   ax = ax, j=j*6, temp_val = temp_lev_value,\n",
    "                                                   _scaler=_scaler,\n",
    "                                                   )\n",
    "                    np.testing.assert_allclose(temp_eval_x[::2], np.arange(1.4,6.3,0.1))\n",
    "\n",
    "                    lw = 3\n",
    "                    if only_below3deg:\n",
    "                        temp_eval_x_sel = temp_eval_x[temp_eval_x<=3]\n",
    "                        _sel_ice_eval_sel = _sel_ice_eval[temp_eval_x<=3]\n",
    "                        colors_fit = cmap(_scaler.transform(temp_eval_x_sel[::2].reshape(-1,1))) # only every decile ... \n",
    "                        for c,f in zip(colors_fit, _sel_ice_eval_sel[::2]):\n",
    "                            ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                                       color=c, zorder=7)\n",
    "                    else:\n",
    "                        colors_fit = cmap(_scaler.transform(temp_eval_x[::2].reshape(-1,1))) # only every decile ... \n",
    "                        for c,f in zip(colors_fit, _sel_ice_eval[::2]):\n",
    "                            ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                                       color=c, zorder=7)\n",
    "\n",
    "                ax.set_xlim([1997, xlim1])\n",
    "                if rgi_reg == 'RGI08':\n",
    "                    ax_fit.set_ylim([-0.5, 120])\n",
    "\n",
    "                ax.set_xticks([2020,2100,2200,2300],[2020,2100,2200,2300])\n",
    "                ax.spines['right'].set_visible(False)\n",
    "                ax.spines['top'].set_visible(False)\n",
    "\n",
    "                fig_fit.tight_layout()\n",
    "                if uncertainty == 'boxplot':\n",
    "                        fig_fit.savefig(f'figures/appendix_fit/lowess_predi_exp_fit_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3.png')\n",
    "\n",
    "            ax_f.text(2290,119, caption_text,  #transform=ax_f.transAxes,\n",
    "                                    verticalalignment=\"top\", bbox=props, horizontalalignment='right',\n",
    "                                    color='grey', alpha=1, fontweight = 'bold', fontsize=18)\n",
    "\n",
    "\n",
    "            if uncertainty == 'boxplot':\n",
    "                qs = [99,103,109,113,120]\n",
    "                q_t = [f' 5%','25%','50%','75%', '95%']\n",
    "                boxplot_manual(qs= qs,\n",
    "                                p = 2300, ax = ax_f, j=0, temp_val = None)\n",
    "                for q,t in zip(qs,q_t):\n",
    "                    ax_f.text(2303, q-1, t, color = 'grey', fontsize=15)\n",
    "                ax_f.set_xticks([2020,2100,2200,2300],[2020,2100,2200,2300])\n",
    "\n",
    "\n",
    "\n",
    "            fig.tight_layout()\n",
    "            # for every of the four large regions ... \n",
    "            if reg == 'Glaciers globally':\n",
    "                reg = 'Glaciers-globally'\n",
    "            if uncertainty== 'boxplot':\n",
    "                if only_below3deg:\n",
    "                    if add_median:\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/png/final_draft_icci_2023_below3deg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.png')\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/svg/final_draft_icci_2023_below3deg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.svg')\n",
    "                    else:\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/png/final_draft_icci_2023_below3deg/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.png')\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/svg/final_draft_icci_2023_below3deg/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.svg')\n",
    "                else:\n",
    "                    if add_median:\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/png/final_draft_icci_2023/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.png')\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/svg/final_draft_icci_2023/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.svg')\n",
    "                    else:\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/png/final_draft_icci_2023/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.png')\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/svg/final_draft_icci_2023/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.svg')\n",
    "\n",
    "\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cba6b0c9",
   "metadata": {},
   "source": [
    "**repeat it for every single RGI region (in separate files)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "2b8697c4",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI16 0.41 0 0.5940125822019444 -0.009331074123102\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI08 0.96 0 1.0382209879829225 0.0437381748806752\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI11 0.51 0 0.4329462619194236 -0.0115070402162595\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI02 0.46 0 0.0863687406451393 -0.0073848595658214\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI18 0.86 0 2.421712130179934 0.0639439491791801\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI06 0.99 0 21.31067898391548 0.075307676741005\n",
      "RGI06 problem\n",
      "RGI06 0.8 0 -1.6090488460476096 -0.2584169616942668\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13-14-15 0.99 0 8.45265502846896 0.0300477867204858\n",
      "RGI13-14-15 0.57 0 -0.0536211913880723 -0.3423459518639633\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI01 0.99 0 24.468128333237928 0.0562669460517852\n",
      "RGI01 0.57 0 0.4378871108982666 -0.0461755604138239\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI04 0.97 0 17.45129269512268 -0.2584482711844496\n",
      "RGI04 0.72 -0.5 -1.453974934319801 -0.1119918149230363\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI17 0.99 0 16.27953236489871 0.3190121418276295\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "global 0.99 0 40.35323776152246 0.1281386443393088\n",
      "global 0.45 0 0.4881256712591804 -0.18261132766947\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI03 problem\n",
      "RGI03 0.86 0 65.2129171468793 -0.379642069470087\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI05 0.99 0 31.33411392714501 0.0360290568379184\n",
      "RGI05 0.64 0 -0.6349166509171063 -0.3024389572831554\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI07 0.76 0 11.748008008139507 -0.0234831019352412\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI09 0.99 0 32.842798576655184 -0.1408007488342377\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI10 0.99 0 1.6764143761648478 0.0695534992579747\n",
      "RGI10 0.52 0 -0.6288899289452097 -0.2626732102131575\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI12 0.7 0 0.8786420268992139 0.0240430667407689\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13 0.99 0 3.4005703198600625 0.0456394446257517\n",
      "RGI13 problem\n",
      "RGI13 0.8 -1.5 -3.487859736807667 -0.290466212611608\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI14 0.99 0 16.40754998175739 0.0381533680331376\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI15 0.78 0 1.3234315964212937 0.0325894930033283\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI19 0.99 0 61.57731305368136 0.0930242333842912\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI16 0.41 0 0.5940125822019444 -0.009331074123102\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI08 0.96 0 1.0382209879829225 0.0437381748806752\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI11 0.51 0 0.4329462619194236 -0.0115070402162595\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI02 0.46 0 0.0863687406451393 -0.0073848595658214\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI18 0.86 0 2.421712130179934 0.0639439491791801\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI06 0.99 0 21.31067898391548 0.075307676741005\n",
      "RGI06 problem\n",
      "RGI06 0.8 0 -1.6090488460476096 -0.2584169616942668\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13-14-15 0.99 0 8.45265502846896 0.0300477867204858\n",
      "RGI13-14-15 0.57 0 -0.0536211913880723 -0.3423459518639633\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI01 0.99 0 24.468128333237928 0.0562669460517852\n",
      "RGI01 0.57 0 0.4378871108982666 -0.0461755604138239\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI04 0.97 0 17.45129269512268 -0.2584482711844496\n",
      "RGI04 0.72 -0.5 -1.453974934319801 -0.1119918149230363\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI17 0.99 0 16.27953236489871 0.3190121418276295\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "global 0.99 0 40.35323776152246 0.1281386443393088\n",
      "global 0.45 0 0.4881256712591804 -0.18261132766947\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI03 problem\n",
      "RGI03 0.86 0 65.2129171468793 -0.379642069470087\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI05 0.99 0 31.33411392714501 0.0360290568379184\n",
      "RGI05 0.64 0 -0.6349166509171063 -0.3024389572831554\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI07 0.76 0 11.748008008139507 -0.0234831019352412\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI09 0.99 0 32.842798576655184 -0.1408007488342377\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI10 0.99 0 1.6764143761648478 0.0695534992579747\n",
      "RGI10 0.52 0 -0.6288899289452097 -0.2626732102131575\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI12 0.7 0 0.8786420268992139 0.0240430667407689\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13 0.99 0 3.4005703198600625 0.0456394446257517\n",
      "RGI13 problem\n",
      "RGI13 0.8 -1.5 -3.487859736807667 -0.290466212611608\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI14 0.99 0 16.40754998175739 0.0381533680331376\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI15 0.78 0 1.3234315964212937 0.0325894930033283\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI19 0.99 0 61.57731305368136 0.0930242333842912\n"
     ]
    }
   ],
   "source": [
    "v = 3 ## COLOR SCALE version ... \n",
    "add_median = True\n",
    "\n",
    "for only_below3deg in [True,False]:\n",
    "\n",
    "    if only_below3deg:\n",
    "        vv = f'v{v}_below3deg'\n",
    "        _scaler = scaler_below3deg\n",
    "        if v == 'm':\n",
    "            cmap = cmap_blue_red_bel3deg\n",
    "        elif v == 1:\n",
    "            cmap = cmap_icci_1\n",
    "        elif v==2:\n",
    "            cmap = cmap_icci_2_sh\n",
    "        elif v==3:\n",
    "            cmap = cmap_icci_below3deg\n",
    "        caption_text = 'colors describe global temperature\\nchange in 2100 from +1.4°C to +3.0°C'\n",
    "\n",
    "    else:\n",
    "        vv = f'v{v}_all'\n",
    "        _scaler = scaler\n",
    "        #cmap = cmap_icci_1 #cmap_icci_2#cmap_blue_red\n",
    "        if v == 'm':\n",
    "            cmap = cmap_blue_red\n",
    "        elif v == 1:\n",
    "            cmap = cmap_icci_1\n",
    "        elif v==2:\n",
    "            cmap = cmap_icci_2\n",
    "        elif v==3:\n",
    "            cmap = cmap_icci_3\n",
    "\n",
    "        caption_text = 'colors describe global temperature\\nchange in 2100 from +1.4°C to +6.3°C'\n",
    "    color_continuous = True\n",
    "    decile_ticks = True\n",
    "    plt.rc('font', size=26)\n",
    "    it = 1\n",
    "    frac_dict = {}\n",
    "    for uncertainty in ['boxplot']: #, 'no_uncertainties']: # 'no_uncertainties',\n",
    "        if uncertainty == 'boxplot':\n",
    "            num_fits = 1000\n",
    "            delta_text = 23\n",
    "            xlim1 = 2327\n",
    "        else:\n",
    "            num_fits = 1000 \n",
    "            delta_text = 4\n",
    "            xlim1 = 2309\n",
    "\n",
    "        for reg in list(d_reg_num_name_icci_reg.keys()):\n",
    "            rgi_regs = d_reg_num_name_icci_reg[reg]\n",
    "            for jj,_rgi_reg in enumerate(rgi_regs): \n",
    "                fig, axs = plt.subplots(1,1, figsize=(12.5,25/3)) \n",
    "                axs = [axs]#.flatten()\n",
    "                #for a in [1,2,3,4,5]:\n",
    "                #    axs[a].remove()\n",
    "                # figure to show caption\n",
    "                ax_f = axs[0] \n",
    "                \n",
    "                ax = axs[0]\n",
    "                list_index_l = []\n",
    "                err_l = []\n",
    "                for m in path_d.keys():\n",
    "                    list_index_l.append(list(pd_cmip_temp_ch_2100.index + '_' + m))\n",
    "                pd_proj_2100_2300 = pd.DataFrame(index=np.concatenate(list_index_l))\n",
    "\n",
    "                if _rgi_reg !='global':\n",
    "                    rgi_reg = f'RGI{_rgi_reg}'\n",
    "                else:\n",
    "                    rgi_reg = _rgi_reg \n",
    "\n",
    "                ####\n",
    "                pd_fit_stats_reg = pd_fit_stats.loc[pd_fit_stats.region == rgi_reg]\n",
    "                pd_fit_stats_reg_lowess = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'lowess_predi']\n",
    "                pd_fit_stats_reg_exp = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'exp']\n",
    "                sel_it = pd_fit_stats_reg_lowess # we don't have it anymore ... .loc[pd_fit_stats_reg_lowess.it ==it]\n",
    "\n",
    "                ######\n",
    "                fig_fit, axs_fit = plt.subplots(2,1, figsize=(15,20)) \n",
    "\n",
    "                for end_year,ls in zip(['2300'],['-']):\n",
    "                    #for temp_lev in colors_temp.keys():\n",
    "\n",
    "                    for cmip in ['CMIP5','CMIP6']:\n",
    "                        pd_cmip = pd_cmip_temp_ch_2100_gcm_until_2300.loc[pd_cmip_temp_ch_2100_gcm_until_2300.ssp != 'ssp534-over']\n",
    "                        pd_cmip = pd_cmip.loc[pd_cmip.cmip == cmip]\n",
    "                        # check that there are temperature values in 2300 available ... \n",
    "                        pd_cmip = pd_cmip.loc[pd_cmip['global_temp_ch_2271-2300_preindustrial'].dropna().index]\n",
    "                        for gcm_scenario in pd_cmip.index:\n",
    "                            df_rel_l = []\n",
    "                            for m in path_d.keys():\n",
    "                                try:\n",
    "                                    gcm, scenario = gcm_scenario.split('_')\n",
    "                                    path = path_d[m]\n",
    "                                    gcm_scenario_m = gcm_scenario + '_' + m\n",
    "                                    n = len(gcm_scenario)\n",
    "                                    if _rgi_reg == '13-14-15':\n",
    "                                        df = 0\n",
    "                                        for sub_rgi_reg in ['RGI13','RGI14','RGI15']:\n",
    "                                            idir = path + f'volume/{cmip}/{end_year}/{sub_rgi_reg}/'\n",
    "                                            _df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                            _df.columns = [g.upper() for g in _df.columns]\n",
    "                                            _df = _df[gcm.upper()]\n",
    "                                            df += _df\n",
    "                                    else:\n",
    "                                        idir = path  + f'volume/{cmip}/{end_year}/{rgi_reg}/'\n",
    "                                        df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                        df.columns = [g.upper() for g in df.columns]\n",
    "                                        df = df[gcm.upper()]\n",
    "                                    df = df.loc[2000:]\n",
    "                                    # first do the rolling than the relative estimates .. \n",
    "                                    df_roll = df.rolling(20, center=True, min_periods=1).mean().dropna()\n",
    "                                    df_rel_roll = 100*df_roll/df_roll.loc[2020]                                    \n",
    "\n",
    "                                    temp_ch = pd_cmip.loc[gcm_scenario, 'global_temp_ch_2071-2100_preindustrial']\n",
    "                                    # if temp_ch < 3:\n",
    "                                    #if only_below3deg:\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, 'global_temp_ch_2071-2100_preindustrial'] = temp_ch\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2100_{rgi_reg}'] = df_rel_roll.loc[2100]\n",
    "\n",
    "                                    color= cmap(_scaler.transform(temp_ch.reshape(-1,1))).flatten()\n",
    "                                    if not add_median:\n",
    "                                        alpha = 0.2\n",
    "                                        ax.plot(df_rel_roll.loc[2020:2300].index, df_rel_roll.loc[2020:2300], lw=2, color=color,ls=ls, alpha = alpha) \n",
    "                                        ax.plot(df_rel_roll.loc[2000:2020].index, df_rel_roll.loc[2000:2020], lw=3, color='black', ls=ls, alpha = alpha)\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2300_{rgi_reg}'] = df_rel_roll.loc[2300]\n",
    "                                    pd_proj_2100_2300.loc[gcm_scenario_m, 'model'] = m\n",
    "\n",
    "                                    df_rel_l.append(df_rel_roll.loc[2000:2300])\n",
    "                                except:\n",
    "                                    pass\n",
    "                            if add_median:\n",
    "                                if len(df_rel_l) > 0:\n",
    "                                    if (not only_below3deg) or (temp_ch<=3):\n",
    "                                        df_rel_med = pd.DataFrame(df_rel_l).T.median(axis=1)\n",
    "                                        ax.plot(df_rel_med.loc[2020:2300].index, df_rel_med.loc[2020:2300], lw=2, color=color,ls=ls, alpha = 0.3) \n",
    "                                        ax.plot(df_rel_med.loc[2000:2020].index, df_rel_med.loc[2000:2020], lw=2, color='black', ls=ls, alpha = 0.3)\n",
    "\n",
    "                ax.set_ylim([0,125])\n",
    "                #if jj in [0,2,4]:\n",
    "                ax.set_ylabel('% of Remaining Ice',# (relative to 2020)',\n",
    "                                  fontweight = 'bold', fontsize=22)\n",
    "                #else:\n",
    "                #    ax.set_ylabel('')\n",
    "                ax.set_xlabel('Year',  fontweight = 'bold', fontsize=22)\n",
    "                handles, labels = ax.get_legend_handles_labels()\n",
    "                for h in handles:\n",
    "                    h.set_linewidth(3)\n",
    "                ax.legend(handles, labels,\n",
    "                              handlelength=1,\n",
    "                              title='Temperature change between 1850-1900 & 2071-2100:', ncol=5)\n",
    "                ax.legend().remove()\n",
    "\n",
    "                ax.set_title(d_reg_num_name_icci[_rgi_reg], fontweight='bold', fontsize=24)\n",
    "\n",
    "                pd_proj_2100_2300 = pd_proj_2100_2300.dropna()\n",
    "                assert len(pd_proj_2100_2300) == 60\n",
    "                for j,sim_year in enumerate([2100,2300]):\n",
    "                    # choose a \"frac\" parameter and select the right fitting values \n",
    "                    frac, _sel = select_frac_w_experts(sel_it, sim_year=sim_year)\n",
    "                    frac_dict[f'{rgi_reg}_{sim_year}'] = frac\n",
    "\n",
    "                    ax_fit = axs_fit[j]\n",
    "                    ax_fit.set_title(f'{d_reg_num_name_icci[_rgi_reg]}:  year {sim_year}', fontweight='bold')\n",
    "\n",
    "                    # get the raw data ...\n",
    "                    x = pd_proj_2100_2300['global_temp_ch_2071-2100_preindustrial']\n",
    "                    y = pd_proj_2100_2300[f'rel_ice_{sim_year}_{rgi_reg}']\n",
    "\n",
    "                    t_min, t_max = x.min(), x.max()\n",
    "                    if sim_year == 2100:\n",
    "                        delta_x = 0.002\n",
    "                    else:\n",
    "                        delta_x = 0.001\n",
    "\n",
    "                    df_quantiles = _sel[['q50','q05','q25','q75','q95']]\n",
    "                    # we just have to clip negative values .... there are still some although we tried to omit them by choosing the best frac ... \n",
    "                    df_quantiles = df_quantiles.where(df_quantiles>=0, 0)\n",
    "                    _sel_ice_eval = df_quantiles['q50'].values\n",
    "                    _sel_ice_eval_bottom = df_quantiles['q05'].values\n",
    "                    _sel_ice_eval_bottom_iqr = df_quantiles['q25'].values\n",
    "                    _sel_ice_eval_top_iqr = df_quantiles['q75'].values\n",
    "                    _sel_ice_eval_top = df_quantiles['q95'].values\n",
    "\n",
    "                    ax_fit.plot(pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].deltaTemp,\n",
    "                             pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].q50, lw=3, color='green', alpha = 0.5, label='exponential fit')\n",
    "                    ax_fit.plot(_sel.deltaTemp,\n",
    "                             _sel_ice_eval, lw=3, color='black', alpha = 1, label=f'lowess, frac={frac}')\n",
    "                    ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom, _sel_ice_eval_top, lw=3, alpha=0.2, color=f'black', label='90% PI')\n",
    "                    ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom_iqr, _sel_ice_eval_top_iqr, lw=3, alpha=0.4, color=f'black', label = '50% PI (IQR)')\n",
    "\n",
    "                    temp_eval_x = _sel.deltaTemp.values\n",
    "                    sns.scatterplot(ax=ax_fit,\n",
    "                                    data = pd_proj_2100_2300, #.loc[j], \n",
    "                                    x='global_temp_ch_2071-2100_preindustrial', y=f'rel_ice_{sim_year}_{rgi_reg}',\n",
    "                                                    style='model',\n",
    "                                    hue='global_temp_ch_2071-2100_preindustrial',\n",
    "                                    palette=cmap,      \n",
    "                                    sizes=[150,150,150], size='model') #, legend=)\n",
    "                    ax_fit.legend(fontsize=16, ncols=2)\n",
    "\n",
    "                    ax_fit.set_xlabel('Temperature change between 1850-1900 & 2071-2100')\n",
    "                    ax_fit.set_ylabel(f'% of Remaining Glacier Ice\\nin {sim_year} (relative to 2020)')\n",
    "\n",
    "                    for j,temp_lev in enumerate(list(colors_temp_values.keys())):\n",
    "\n",
    "                        # for the annotation in the main plot ... \n",
    "                        if (temp_lev != '>4.0°C') and (temp_lev != '<=+4.0°C'):\n",
    "\n",
    "                            temp_lev_value = colors_temp_values[temp_lev]\n",
    "                            color= cmap(_scaler.transform(np.array([temp_lev_value]).reshape(-1,1))).flatten()\n",
    "                            idx_nearest = np.abs(temp_eval_x - temp_lev_value).argmin()\n",
    "                            if j >0:\n",
    "                                # check the previous fit, if they overlap, need to shift the text\n",
    "                                previous_fit = fit.copy()\n",
    "                            else:\n",
    "                                previous_fit = 10\n",
    "\n",
    "\n",
    "                            if previous_fit <4:\n",
    "                                shift = previous_fit-3.5\n",
    "                            else:\n",
    "                                shift = 0\n",
    "\n",
    "                            fit = _sel_ice_eval[idx_nearest]\n",
    "\n",
    "                            if sim_year == 2300:\n",
    "                                loc_2300_stats = 2300 +4 \n",
    "                            else:\n",
    "                                loc_2300_stats = 2100\n",
    "                            #if temp_lev_value>1.5:\n",
    "                            #    ax.vlines(loc_2300_stats,y_fitted[0], y_fitted[1], color=colors_temp[temp_lev], lw=15, alpha = 0.4)\n",
    "                            ax.hlines(fit,loc_2300_stats-3, loc_2300_stats+3, lw=8, color=color, zorder=10)\n",
    "\n",
    "                            if sim_year == 2300:\n",
    "                                bbox=dict(facecolor='white', edgecolor='white', boxstyle='round', alpha = 0.8)\n",
    "                                ax.annotate(temp_lev[2:], xy=(2300, fit),\n",
    "                                            xytext=(loc_2300_stats+delta_text, fit+shift),\n",
    "                                                        verticalalignment=\"center\",\n",
    "                                                        color=color, alpha=1, fontweight = 'bold', fontsize=18)\n",
    "                                _sel_temp_lev_qs = np.array([ _sel_ice_eval_bottom[idx_nearest],\n",
    "                                                             _sel_ice_eval_bottom_iqr[idx_nearest], _sel_ice_eval[idx_nearest],\n",
    "                                                             _sel_ice_eval_top_iqr[idx_nearest], _sel_ice_eval_top[idx_nearest]])\n",
    "                                if uncertainty == 'boxplot':\n",
    "                                    boxplot_manual(qs= _sel_temp_lev_qs,\n",
    "                                                   p = loc_2300_stats+7,\n",
    "                                                   ax = ax, j=j*6, temp_val = temp_lev_value,\n",
    "                                                   _scaler=_scaler,\n",
    "                                                   )\n",
    "                    np.testing.assert_allclose(temp_eval_x[::2], np.arange(1.4,6.3,0.1))\n",
    "\n",
    "                    lw = 3\n",
    "                    if only_below3deg:\n",
    "                        temp_eval_x_sel = temp_eval_x[temp_eval_x<=3]\n",
    "                        _sel_ice_eval_sel = _sel_ice_eval[temp_eval_x<=3]\n",
    "                        colors_fit = cmap(_scaler.transform(temp_eval_x_sel[::2].reshape(-1,1))) # only every decile ... \n",
    "                        for c,f in zip(colors_fit, _sel_ice_eval_sel[::2]):\n",
    "                            ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                                       color=c, zorder=7)\n",
    "                    else:\n",
    "                        colors_fit = cmap(_scaler.transform(temp_eval_x[::2].reshape(-1,1))) # only every decile ... \n",
    "                        for c,f in zip(colors_fit, _sel_ice_eval[::2]):\n",
    "                            ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                                       color=c, zorder=7)\n",
    "\n",
    "                ax.set_xlim([1997, xlim1])\n",
    "                if rgi_reg == 'RGI08':\n",
    "                    ax_fit.set_ylim([-0.5, 120])\n",
    "\n",
    "                ax.set_xticks([2020,2100,2200,2300],[2020,2100,2200,2300])\n",
    "                ax.spines['right'].set_visible(False)\n",
    "                ax.spines['top'].set_visible(False)\n",
    "\n",
    "                fig_fit.tight_layout()\n",
    "                if uncertainty == 'boxplot':\n",
    "                        fig_fit.savefig(f'figures/appendix_fit/lowess_predi_exp_fit_oggm_glogem_pygem_temp_levels_{rgi_reg}_v_final.png')\n",
    "                \n",
    "                fig.tight_layout()\n",
    "\n",
    "                if uncertainty== 'boxplot':\n",
    "                    if add_median:\n",
    "                        #fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/png/final_draft_icci_2023_below3deg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.png')\n",
    "                        #fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/svg/final_draft_icci_2023_below3deg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.svg')\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/png/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_{vv}_final_{uncertainty}.png')\n",
    "                        fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/svg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_{vv}_final_{uncertainty}.svg')\n",
    "\n",
    "                plt.close()\n",
    "                plt.close()\n",
    "                plt.close()\n",
    "                plt.close()\n",
    "                plt.close()\n",
    "                plt.close()\n",
    "                plt.close()\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "87a03151",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3b0efdad",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "9e93dd0a",
   "metadata": {},
   "source": [
    "## Variant: Only below 4.5 degree celsius"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "e32e0e5b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.427504187860568 4.35665250612107\n"
     ]
    }
   ],
   "source": [
    "pd_cmip_sel_below45deg = pd_cmip_temp_ch_2100_gcm_until_2300.loc[pd_cmip_temp_ch_2100_gcm_until_2300['global_temp_ch_2071-2100_preindustrial']<=4.5]\n",
    "data_temp_bel45deg = pd_cmip_sel_below45deg['global_temp_ch_2071-2100_preindustrial'].values.reshape(1,-1).T\n",
    "scaler_below45deg = MinMaxScaler()\n",
    "scaler_below45deg.fit(data_temp_bel45deg)\n",
    "t_min_bel45deg, t_max_bel45deg = pd_cmip_sel_below45deg['global_temp_ch_2071-2100_preindustrial'].min(), pd_cmip_sel_below45deg['global_temp_ch_2071-2100_preindustrial'].max()\n",
    "print(t_min_bel45deg, t_max_bel45deg)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a69dffa2",
   "metadata": {},
   "source": [
    "**amount of climate experiments**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "7b196804",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "19"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(pd_cmip_sel_below45deg)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "id": "1a1954ef",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI16 0.41 0 0.5940125822019444 -0.009331074123102\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI08 0.96 0 1.0382209879829225 0.0437381748806752\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI11 0.51 0 0.4329462619194236 -0.0115070402162595\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI02 0.46 0 0.0863687406451393 -0.0073848595658214\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI18 0.86 0 2.421712130179934 0.0639439491791801\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI06 0.99 0 21.31067898391548 0.075307676741005\n",
      "RGI06 problem\n",
      "RGI06 0.8 0 -1.6090488460476096 -0.2584169616942668\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13-14-15 0.99 0 8.45265502846896 0.0300477867204858\n",
      "RGI13-14-15 0.57 0 -0.0536211913880723 -0.3423459518639633\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI01 0.99 0 24.468128333237928 0.0562669460517852\n",
      "RGI01 0.57 0 0.4378871108982666 -0.0461755604138239\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI04 0.97 0 17.45129269512268 -0.2584482711844496\n",
      "RGI04 0.72 -0.5 -1.453974934319801 -0.1119918149230363\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI17 0.99 0 16.27953236489871 0.3190121418276295\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "global 0.99 0 40.35323776152246 0.1281386443393088\n",
      "global 0.45 0 0.4881256712591804 -0.18261132766947\n"
     ]
    }
   ],
   "source": [
    "v = 3\n",
    "vv = f'v{v}_below45deg'\n",
    "add_median = True\n",
    "only_below45deg = True\n",
    "#################################################\n",
    "_scaler = scaler_below45deg  \n",
    "if v==3:\n",
    "    cmap = cmap_icci_3\n",
    "elif v==4:\n",
    "    cmap = cmap_icci_below3deg # it was designed for below 3, but should also work below 4.5\n",
    "\n",
    "caption_text = 'colors describe global temperature\\nchange in 2100 from +1.4°C to +4.5°C'\n",
    "################################################################\n",
    "color_continuous = True\n",
    "decile_ticks = True\n",
    "plt.rc('font', size=26)\n",
    "it = 1\n",
    "frac_dict = {}\n",
    "for uncertainty in [ 'boxplot']: #, 'no_uncertainties']: # 'no_uncertainties',\n",
    "    if uncertainty == 'boxplot':\n",
    "        num_fits = 1000\n",
    "        delta_text = 23\n",
    "        xlim1 = 2327\n",
    "    else:\n",
    "        num_fits = 1000 \n",
    "        delta_text = 4\n",
    "        xlim1 = 2309\n",
    "\n",
    "    for reg in list(d_reg_num_name_icci_reg.keys()):\n",
    "        rgi_regs = d_reg_num_name_icci_reg[reg]\n",
    "        \n",
    "        len_sub_p = len(rgi_regs)\n",
    "        props = dict(boxstyle='round', facecolor='white', alpha=0.5)\n",
    "        if (reg == 'Glaciers globally') or (reg == 'low-latitudes'):\n",
    "            fig, axs = plt.subplots(1,1, figsize=(12.5,25/3)) \n",
    "            axs = [axs]#.flatten()\n",
    "            #for a in [1,2,3,4,5]:\n",
    "            #    axs[a].remove()\n",
    "            # figure to show caption \n",
    "            ax_f = axs[0] \n",
    "           \n",
    "        elif reg == 'mid-latitudes':\n",
    "            fig, axs = plt.subplots(2,2, figsize=(25,25*2/3))\n",
    "            axs = axs.flatten()\n",
    "            #axs[4].remove()\n",
    "            #axs[5].remove()\n",
    "            ax_f = axs[2]\n",
    "        elif reg == 'polar-hma':\n",
    "            fig, axs = plt.subplots(3,2, figsize=(25,25))\n",
    "            axs = axs.flatten()\n",
    "            ax_f = axs[4]\n",
    "            #ax_f.spines['left'].set_visible(False)\n",
    "            #ax_f.spines['bottom'].set_visible(False)\n",
    "            axs[5].remove()\n",
    "            \n",
    "        else:\n",
    "            print('sth. wrong')\n",
    "\n",
    "        for jj,_rgi_reg in enumerate(rgi_regs): \n",
    "            ax = axs[jj]\n",
    "            list_index_l = []\n",
    "            err_l = []\n",
    "            for m in path_d.keys():\n",
    "                list_index_l.append(list(pd_cmip_temp_ch_2100.index + '_' + m))\n",
    "            pd_proj_2100_2300 = pd.DataFrame(index=np.concatenate(list_index_l))\n",
    "\n",
    "            if _rgi_reg !='global':\n",
    "                rgi_reg = f'RGI{_rgi_reg}'\n",
    "            else:\n",
    "                rgi_reg = _rgi_reg \n",
    "                \n",
    "            ####\n",
    "            pd_fit_stats_reg = pd_fit_stats.loc[pd_fit_stats.region == rgi_reg]\n",
    "            pd_fit_stats_reg_lowess = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'lowess_predi']\n",
    "            pd_fit_stats_reg_exp = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'exp']\n",
    "            sel_it = pd_fit_stats_reg_lowess # we don't have it anymore ... .loc[pd_fit_stats_reg_lowess.it ==it]\n",
    "\n",
    "            ######\n",
    "            fig_fit, axs_fit = plt.subplots(2,1, figsize=(15,20)) \n",
    "            \n",
    "            for end_year,ls in zip(['2300'],['-']):\n",
    "                #for temp_lev in colors_temp.keys():\n",
    "\n",
    "                for cmip in ['CMIP5','CMIP6']:\n",
    "                    pd_cmip = pd_cmip_temp_ch_2100_gcm_until_2300.loc[pd_cmip_temp_ch_2100_gcm_until_2300.ssp != 'ssp534-over']\n",
    "                    pd_cmip = pd_cmip.loc[pd_cmip.cmip == cmip]\n",
    "                    # check that there are temperature values in 2300 available ... \n",
    "                    pd_cmip = pd_cmip.loc[pd_cmip['global_temp_ch_2271-2300_preindustrial'].dropna().index]\n",
    "                    for gcm_scenario in pd_cmip.index:\n",
    "                        df_rel_l = []\n",
    "                        for m in path_d.keys():\n",
    "                            try:\n",
    "                                gcm, scenario = gcm_scenario.split('_')\n",
    "                                path = path_d[m]\n",
    "                                gcm_scenario_m = gcm_scenario + '_' + m\n",
    "                                n = len(gcm_scenario)\n",
    "                                if _rgi_reg == '13-14-15':\n",
    "                                    df = 0\n",
    "                                    for sub_rgi_reg in ['RGI13','RGI14','RGI15']:\n",
    "                                        idir = path + f'volume/{cmip}/{end_year}/{sub_rgi_reg}/'\n",
    "                                        _df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                        _df.columns = [g.upper() for g in _df.columns]\n",
    "                                        _df = _df[gcm.upper()]\n",
    "                                        df += _df\n",
    "                                else:\n",
    "                                    idir = path  + f'volume/{cmip}/{end_year}/{rgi_reg}/'\n",
    "                                    df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                    df.columns = [g.upper() for g in df.columns]\n",
    "                                    df = df[gcm.upper()]\n",
    "                                df = df.loc[2000:]\n",
    "                                # first do the rolling than the relative estimates .. \n",
    "                                df_roll = df.rolling(20, center=True, min_periods=1).mean().dropna()\n",
    "                                df_rel_roll = 100*df_roll/df_roll.loc[2020]                                    \n",
    "\n",
    "                                temp_ch = pd_cmip.loc[gcm_scenario, 'global_temp_ch_2071-2100_preindustrial']\n",
    "                                # if temp_ch < 3:\n",
    "                                #if only_below3deg:\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, 'global_temp_ch_2071-2100_preindustrial'] = temp_ch\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2100_{rgi_reg}'] = df_rel_roll.loc[2100]\n",
    "\n",
    "                                color= cmap(_scaler.transform(temp_ch.reshape(-1,1))).flatten()\n",
    "                                if not add_median:\n",
    "                                    alpha = 0.2\n",
    "                                    ax.plot(df_rel_roll.loc[2020:2300].index, df_rel_roll.loc[2020:2300], lw=2, color=color,ls=ls, alpha = alpha) \n",
    "                                    ax.plot(df_rel_roll.loc[2000:2020].index, df_rel_roll.loc[2000:2020], lw=3, color='black', ls=ls, alpha = alpha)\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2300_{rgi_reg}'] = df_rel_roll.loc[2300]\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, 'model'] = m\n",
    "\n",
    "                                df_rel_l.append(df_rel_roll.loc[2000:2300])\n",
    "                            except:\n",
    "                                pass\n",
    "                        if add_median:\n",
    "                            if len(df_rel_l) > 0:\n",
    "                                if (not only_below45deg) or (temp_ch<=4.5):\n",
    "                                    df_rel_med = pd.DataFrame(df_rel_l).T.median(axis=1)\n",
    "                                    ax.plot(df_rel_med.loc[2020:2300].index, df_rel_med.loc[2020:2300], lw=2, color=color,ls=ls, alpha = 0.3) \n",
    "                                    ax.plot(df_rel_med.loc[2000:2020].index, df_rel_med.loc[2000:2020], lw=2, color='black', ls=ls, alpha = 0.3)\n",
    "\n",
    "            ax.set_ylim([0,125])\n",
    "            if jj in [0,2,4]:\n",
    "                ax.set_ylabel('% of Remaining Ice',# (relative to 2020)',\n",
    "                              fontweight = 'bold', fontsize=22)\n",
    "            else:\n",
    "                ax.set_ylabel('')\n",
    "            ax.set_xlabel('Year',  fontweight = 'bold', fontsize=22)\n",
    "            handles, labels = ax.get_legend_handles_labels()\n",
    "            for h in handles:\n",
    "                h.set_linewidth(3)\n",
    "            ax.legend(handles, labels,\n",
    "                          handlelength=1,\n",
    "                          title='Temperature change between 1850-1900 & 2071-2100:', ncol=5)\n",
    "            ax.legend().remove()\n",
    "\n",
    "            ax.set_title(d_reg_num_name_icci[_rgi_reg], fontweight='bold', fontsize=24)\n",
    "\n",
    "            pd_proj_2100_2300 = pd_proj_2100_2300.dropna()\n",
    "            assert len(pd_proj_2100_2300) == 60\n",
    "            for j,sim_year in enumerate([2100,2300]):\n",
    "                # choose a \"frac\" parameter and select the right fitting values \n",
    "                frac, _sel = select_frac_w_experts(sel_it, sim_year=sim_year)\n",
    "                frac_dict[f'{rgi_reg}_{sim_year}'] = frac\n",
    "                \n",
    "                ax_fit = axs_fit[j]\n",
    "                ax_fit.set_title(f'{d_reg_num_name_icci[_rgi_reg]}:  year {sim_year}', fontweight='bold')\n",
    "\n",
    "                # get the raw data ...\n",
    "                x = pd_proj_2100_2300['global_temp_ch_2071-2100_preindustrial']\n",
    "                y = pd_proj_2100_2300[f'rel_ice_{sim_year}_{rgi_reg}']\n",
    "                \n",
    "                t_min, t_max = x.min(), x.max()\n",
    "                if sim_year == 2100:\n",
    "                    delta_x = 0.002\n",
    "                else:\n",
    "                    delta_x = 0.001\n",
    "\n",
    "                df_quantiles = _sel[['q50','q05','q25','q75','q95']]\n",
    "                # we just have to clip negative values .... there are still some although we tried to omit them by choosing the best frac ... \n",
    "                df_quantiles = df_quantiles.where(df_quantiles>=0, 0)\n",
    "                _sel_ice_eval = df_quantiles['q50'].values\n",
    "                _sel_ice_eval_bottom = df_quantiles['q05'].values\n",
    "                _sel_ice_eval_bottom_iqr = df_quantiles['q25'].values\n",
    "                _sel_ice_eval_top_iqr = df_quantiles['q75'].values\n",
    "                _sel_ice_eval_top = df_quantiles['q95'].values\n",
    "\n",
    "                ax_fit.plot(pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].deltaTemp,\n",
    "                         pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].q50, lw=3, color='green', alpha = 0.5, label='exponential fit')\n",
    "                ax_fit.plot(_sel.deltaTemp,\n",
    "                         _sel_ice_eval, lw=3, color='black', alpha = 1, label=f'lowess, frac={frac}')\n",
    "                ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom, _sel_ice_eval_top, lw=3, alpha=0.2, color=f'black', label='90% PI')\n",
    "                ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom_iqr, _sel_ice_eval_top_iqr, lw=3, alpha=0.4, color=f'black', label = '50% PI (IQR)')\n",
    "                \n",
    "                temp_eval_x = _sel.deltaTemp.values\n",
    "                sns.scatterplot(ax=ax_fit,\n",
    "                                data = pd_proj_2100_2300, #.loc[j], \n",
    "                                x='global_temp_ch_2071-2100_preindustrial', y=f'rel_ice_{sim_year}_{rgi_reg}',\n",
    "                                                style='model',\n",
    "                                hue='global_temp_ch_2071-2100_preindustrial',\n",
    "                                palette=cmap,      \n",
    "                                sizes=[150,150,150], size='model') #, legend=)\n",
    "                ax_fit.legend(fontsize=16, ncols=2)\n",
    "\n",
    "                ax_fit.set_xlabel('Temperature change between 1850-1900 & 2071-2100')\n",
    "                ax_fit.set_ylabel(f'% of Remaining Glacier Ice\\nin {sim_year} (relative to 2020)')\n",
    "\n",
    "                for j,temp_lev in enumerate(list(colors_temp_values.keys())):\n",
    "\n",
    "                    # for the annotation in the main plot ... \n",
    "                    if (temp_lev != '>4.0°C') and (temp_lev != '<=+4.0°C'):\n",
    "\n",
    "                        temp_lev_value = colors_temp_values[temp_lev]\n",
    "                        color= cmap(_scaler.transform(np.array([temp_lev_value]).reshape(-1,1))).flatten()\n",
    "                        idx_nearest = np.abs(temp_eval_x - temp_lev_value).argmin()\n",
    "                        if j >0:\n",
    "                            # check the previous fit, if they overlap, need to shift the text\n",
    "                            previous_fit = fit.copy()\n",
    "                        else:\n",
    "                            previous_fit = 10\n",
    "                            \n",
    "                            \n",
    "                        if previous_fit <4:\n",
    "                            shift = previous_fit-3.5\n",
    "                        else:\n",
    "                            shift = 0\n",
    "                        \n",
    "                        fit = _sel_ice_eval[idx_nearest]\n",
    "\n",
    "\n",
    "\n",
    "                        if sim_year == 2300:\n",
    "                            loc_2300_stats = 2300 +4 \n",
    "                        else:\n",
    "                            loc_2300_stats = 2100\n",
    "                        #if temp_lev_value>1.5:\n",
    "                        #    ax.vlines(loc_2300_stats,y_fitted[0], y_fitted[1], color=colors_temp[temp_lev], lw=15, alpha = 0.4)\n",
    "                        ax.hlines(fit,loc_2300_stats-3, loc_2300_stats+3, lw=8, color=color, zorder=10)\n",
    "\n",
    "                        if sim_year == 2300:\n",
    "                            bbox=dict(facecolor='white', edgecolor='white', boxstyle='round', alpha = 0.8)\n",
    "                            ax.annotate(temp_lev[2:], xy=(2300, fit),\n",
    "                                        xytext=(loc_2300_stats+delta_text, fit+shift),\n",
    "                                                    verticalalignment=\"center\",\n",
    "                                                    color=color, alpha=1, fontweight = 'bold', fontsize=18)\n",
    "                            _sel_temp_lev_qs = np.array([ _sel_ice_eval_bottom[idx_nearest],\n",
    "                                                         _sel_ice_eval_bottom_iqr[idx_nearest], _sel_ice_eval[idx_nearest],\n",
    "                                                         _sel_ice_eval_top_iqr[idx_nearest], _sel_ice_eval_top[idx_nearest]])\n",
    "                            if uncertainty == 'boxplot':\n",
    "                                boxplot_manual(qs= _sel_temp_lev_qs,\n",
    "                                               p = loc_2300_stats+7,\n",
    "                                               ax = ax, j=j*6, temp_val = temp_lev_value,\n",
    "                                               _scaler=_scaler,\n",
    "                                               )\n",
    "                np.testing.assert_allclose(temp_eval_x[::2], np.arange(1.4,6.3,0.1))\n",
    "                \n",
    "                lw = 3\n",
    "                if only_below45deg:\n",
    "                    temp_eval_x_sel = temp_eval_x[temp_eval_x<=4.5]\n",
    "                    _sel_ice_eval_sel = _sel_ice_eval[temp_eval_x<=4.5]\n",
    "                    colors_fit = cmap(_scaler.transform(temp_eval_x_sel[::2].reshape(-1,1))) # only every decile ... \n",
    "                    for c,f in zip(colors_fit, _sel_ice_eval_sel[::2]):\n",
    "                        ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                                   color=c, zorder=7)\n",
    "                else:\n",
    "                    colors_fit = cmap(_scaler.transform(temp_eval_x[::2].reshape(-1,1))) # only every decile ... \n",
    "                    for c,f in zip(colors_fit, _sel_ice_eval[::2]):\n",
    "                        ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                                   color=c, zorder=7)\n",
    "\n",
    "            ax.set_xlim([1997, xlim1])\n",
    "            if rgi_reg == 'RGI08':\n",
    "                ax_fit.set_ylim([-0.5, 120])\n",
    "\n",
    "            ax.set_xticks([2020,2100,2200,2300],[2020,2100,2200,2300])\n",
    "            ax.spines['right'].set_visible(False)\n",
    "            ax.spines['top'].set_visible(False)\n",
    "            \n",
    "            fig_fit.tight_layout()\n",
    "            if uncertainty == 'boxplot':\n",
    "                    fig_fit.savefig(f'figures/appendix_fit/lowess_predi_exp_fit_oggm_glogem_pygem_temp_levels_{rgi_reg}_v3.png')\n",
    "        \n",
    "        ax_f.text(2290,119, caption_text,  #transform=ax_f.transAxes,\n",
    "                                verticalalignment=\"top\", bbox=props, horizontalalignment='right',\n",
    "                                color='grey', alpha=1, fontweight = 'bold', fontsize=18)\n",
    "        \n",
    "\n",
    "        if uncertainty == 'boxplot':\n",
    "            qs = [99,103,109,113,120]\n",
    "            q_t = [f' 5%','25%','50%','75%', '95%']\n",
    "            boxplot_manual(qs= qs,\n",
    "                            p = 2300, ax = ax_f, j=0, temp_val = None)\n",
    "            for q,t in zip(qs,q_t):\n",
    "                ax_f.text(2303, q-1, t, color = 'grey', fontsize=15)\n",
    "            ax_f.set_xticks([2020,2100,2200,2300],[2020,2100,2200,2300])\n",
    "\n",
    "        fig.tight_layout()\n",
    "        # for every of the four large regions ... \n",
    "        if reg == 'Glaciers globally':\n",
    "            reg = 'Glaciers-globally'\n",
    "        if uncertainty== 'boxplot':\n",
    "            if only_below45deg:\n",
    "                if add_median:\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/png/final_draft_icci_2023_below45deg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.png')\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/svg/final_draft_icci_2023_below45deg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.svg')\n",
    "                else:\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/png/final_draft_icci_2023_below45deg/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.png')\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/svg/final_draft_icci_2023_below45deg/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.svg')\n",
    "            else:\n",
    "                if add_median:\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/png/final_draft_icci_2023/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.png')\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/svg/final_draft_icci_2023/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.svg')\n",
    "                else:\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/png/final_draft_icci_2023/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.png')\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/all_glac_models/svg/final_draft_icci_2023/icci_report_2023_lowess_predi_oggm_glogem_pygem_temp_levels_{reg}_{uncertainty}_{vv}.svg')\n",
    "                \n",
    "\n",
    "        plt.close()\n",
    "        plt.close()\n",
    "        plt.close()\n",
    "        plt.close()\n",
    "        plt.close()\n",
    "        plt.close()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f79ee91a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "d74a10b8",
   "metadata": {},
   "source": [
    "**repeat for every single region**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "4afd16b0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI16 0.41 0 0.5940125822019444 -0.009331074123102\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI08 0.96 0 1.0382209879829225 0.0437381748806752\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI11 0.51 0 0.4329462619194236 -0.0115070402162595\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI02 0.46 0 0.0863687406451393 -0.0073848595658214\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI18 0.86 0 2.421712130179934 0.0639439491791801\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI06 0.99 0 21.31067898391548 0.075307676741005\n",
      "RGI06 problem\n",
      "RGI06 0.8 0 -1.6090488460476096 -0.2584169616942668\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13-14-15 0.99 0 8.45265502846896 0.0300477867204858\n",
      "RGI13-14-15 0.57 0 -0.0536211913880723 -0.3423459518639633\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI01 0.99 0 24.468128333237928 0.0562669460517852\n",
      "RGI01 0.57 0 0.4378871108982666 -0.0461755604138239\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI04 0.97 0 17.45129269512268 -0.2584482711844496\n",
      "RGI04 0.72 -0.5 -1.453974934319801 -0.1119918149230363\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI17 0.99 0 16.27953236489871 0.3190121418276295\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "global 0.99 0 40.35323776152246 0.1281386443393088\n",
      "global 0.45 0 0.4881256712591804 -0.18261132766947\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI03 problem\n",
      "RGI03 0.86 0 65.2129171468793 -0.379642069470087\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI05 0.99 0 31.33411392714501 0.0360290568379184\n",
      "RGI05 0.64 0 -0.6349166509171063 -0.3024389572831554\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI07 0.76 0 11.748008008139507 -0.0234831019352412\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI09 0.99 0 32.842798576655184 -0.1408007488342377\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI10 0.99 0 1.6764143761648478 0.0695534992579747\n",
      "RGI10 0.52 0 -0.6288899289452097 -0.2626732102131575\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI12 0.7 0 0.8786420268992139 0.0240430667407689\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI13 0.99 0 3.4005703198600625 0.0456394446257517\n",
      "RGI13 problem\n",
      "RGI13 0.8 -1.5 -3.487859736807667 -0.290466212611608\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI14 0.99 0 16.40754998175739 0.0381533680331376\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI15 0.78 0 1.3234315964212937 0.0325894930033283\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No artists with labels found to put in legend.  Note that artists whose label start with an underscore are ignored when legend() is called with no argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RGI19 0.99 0 61.57731305368136 0.0930242333842912\n"
     ]
    }
   ],
   "source": [
    "v = 3\n",
    "vv = f'v{v}_below45deg'\n",
    "add_median = True\n",
    "only_below45deg = True\n",
    "#################################################\n",
    "_scaler = scaler_below45deg  \n",
    "if v==3:\n",
    "    cmap = cmap_icci_3\n",
    "elif v==4:\n",
    "    cmap = cmap_icci_below3deg # it was designed for below 3, but should also work below 4.5\n",
    "\n",
    "caption_text = 'colors describe global temperature\\nchange in 2100 from +1.4°C to +4.5°C'\n",
    "props = dict(boxstyle='round', facecolor='white', alpha=0.5)\n",
    "################################################################\n",
    "color_continuous = True\n",
    "decile_ticks = True\n",
    "plt.rc('font', size=26)\n",
    "it = 1\n",
    "frac_dict = {}\n",
    "for uncertainty in [ 'boxplot']: #, 'no_uncertainties']: # 'no_uncertainties',\n",
    "    if uncertainty == 'boxplot':\n",
    "        num_fits = 1000\n",
    "        delta_text = 23\n",
    "        xlim1 = 2327\n",
    "    else:\n",
    "        num_fits = 1000 \n",
    "        delta_text = 4\n",
    "        xlim1 = 2309\n",
    "\n",
    "    for reg in list(d_reg_num_name_icci_reg.keys()):\n",
    "        rgi_regs = d_reg_num_name_icci_reg[reg]\n",
    "        \n",
    "        len_sub_p = len(rgi_regs)\n",
    "\n",
    "        for jj,_rgi_reg in enumerate(rgi_regs): \n",
    "            fig, axs = plt.subplots(1,1, figsize=(12.5,25/3)) \n",
    "            axs = [axs]\n",
    "            # figure to show caption\n",
    "            ax_f = axs[0] \n",
    "            ax = axs[0]\n",
    "            \n",
    "            \n",
    "            list_index_l = []\n",
    "            err_l = []\n",
    "            for m in path_d.keys():\n",
    "                list_index_l.append(list(pd_cmip_temp_ch_2100.index + '_' + m))\n",
    "            pd_proj_2100_2300 = pd.DataFrame(index=np.concatenate(list_index_l))\n",
    "\n",
    "            if _rgi_reg !='global':\n",
    "                rgi_reg = f'RGI{_rgi_reg}'\n",
    "            else:\n",
    "                rgi_reg = _rgi_reg \n",
    "                \n",
    "            ####\n",
    "            pd_fit_stats_reg = pd_fit_stats.loc[pd_fit_stats.region == rgi_reg]\n",
    "            pd_fit_stats_reg_lowess = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'lowess_predi']\n",
    "            pd_fit_stats_reg_exp = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'exp']\n",
    "            sel_it = pd_fit_stats_reg_lowess # we don't have it anymore ... .loc[pd_fit_stats_reg_lowess.it ==it]\n",
    "\n",
    "            ######\n",
    "            fig_fit, axs_fit = plt.subplots(2,1, figsize=(15,20)) \n",
    "            \n",
    "            for end_year,ls in zip(['2300'],['-']):\n",
    "                #for temp_lev in colors_temp.keys():\n",
    "\n",
    "                for cmip in ['CMIP5','CMIP6']:\n",
    "                    pd_cmip = pd_cmip_temp_ch_2100_gcm_until_2300.loc[pd_cmip_temp_ch_2100_gcm_until_2300.ssp != 'ssp534-over']\n",
    "                    pd_cmip = pd_cmip.loc[pd_cmip.cmip == cmip]\n",
    "                    # check that there are temperature values in 2300 available ... \n",
    "                    pd_cmip = pd_cmip.loc[pd_cmip['global_temp_ch_2271-2300_preindustrial'].dropna().index]\n",
    "                    for gcm_scenario in pd_cmip.index:\n",
    "                        df_rel_l = []\n",
    "                        for m in path_d.keys():\n",
    "                            try:\n",
    "                                gcm, scenario = gcm_scenario.split('_')\n",
    "                                path = path_d[m]\n",
    "                                gcm_scenario_m = gcm_scenario + '_' + m\n",
    "                                n = len(gcm_scenario)\n",
    "                                if _rgi_reg == '13-14-15':\n",
    "                                    df = 0\n",
    "                                    for sub_rgi_reg in ['RGI13','RGI14','RGI15']:\n",
    "                                        idir = path + f'volume/{cmip}/{end_year}/{sub_rgi_reg}/'\n",
    "                                        _df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                        _df.columns = [g.upper() for g in _df.columns]\n",
    "                                        _df = _df[gcm.upper()]\n",
    "                                        df += _df\n",
    "                                else:\n",
    "                                    idir = path  + f'volume/{cmip}/{end_year}/{rgi_reg}/'\n",
    "                                    df = pd.read_csv(idir + f'{scenario}.csv', index_col=0)\n",
    "                                    df.columns = [g.upper() for g in df.columns]\n",
    "                                    df = df[gcm.upper()]\n",
    "                                df = df.loc[2000:]\n",
    "                                # first do the rolling than the relative estimates .. \n",
    "                                df_roll = df.rolling(20, center=True, min_periods=1).mean().dropna()\n",
    "                                df_rel_roll = 100*df_roll/df_roll.loc[2020]                                    \n",
    "\n",
    "                                temp_ch = pd_cmip.loc[gcm_scenario, 'global_temp_ch_2071-2100_preindustrial']\n",
    "                                # if temp_ch < 3:\n",
    "                                #if only_below3deg:\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, 'global_temp_ch_2071-2100_preindustrial'] = temp_ch\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2100_{rgi_reg}'] = df_rel_roll.loc[2100]\n",
    "\n",
    "                                color= cmap(_scaler.transform(temp_ch.reshape(-1,1))).flatten()\n",
    "                                if not add_median:\n",
    "                                    alpha = 0.2\n",
    "                                    ax.plot(df_rel_roll.loc[2020:2300].index, df_rel_roll.loc[2020:2300], lw=2, color=color,ls=ls, alpha = alpha) \n",
    "                                    ax.plot(df_rel_roll.loc[2000:2020].index, df_rel_roll.loc[2000:2020], lw=3, color='black', ls=ls, alpha = alpha)\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, f'rel_ice_2300_{rgi_reg}'] = df_rel_roll.loc[2300]\n",
    "                                pd_proj_2100_2300.loc[gcm_scenario_m, 'model'] = m\n",
    "\n",
    "                                df_rel_l.append(df_rel_roll.loc[2000:2300])\n",
    "                            except:\n",
    "                                pass\n",
    "                        if add_median:\n",
    "                            if len(df_rel_l) > 0:\n",
    "                                if (not only_below45deg) or (temp_ch<=4.5):\n",
    "                                    df_rel_med = pd.DataFrame(df_rel_l).T.median(axis=1)\n",
    "                                    ax.plot(df_rel_med.loc[2020:2300].index, df_rel_med.loc[2020:2300], lw=2, color=color,ls=ls, alpha = 0.3) \n",
    "                                    ax.plot(df_rel_med.loc[2000:2020].index, df_rel_med.loc[2000:2020], lw=2, color='black', ls=ls, alpha = 0.3)\n",
    "\n",
    "            ax.set_ylim([0,125])\n",
    "            ax.set_ylabel('% of Remaining Ice',# (relative to 2020)',\n",
    "                          fontweight = 'bold', fontsize=22)\n",
    "            \n",
    "            ax.set_xlabel('Year',  fontweight = 'bold', fontsize=22)\n",
    "            handles, labels = ax.get_legend_handles_labels()\n",
    "            for h in handles:\n",
    "                h.set_linewidth(3)\n",
    "            ax.legend(handles, labels,\n",
    "                          handlelength=1,\n",
    "                          title='Temperature change between 1850-1900 & 2071-2100:', ncol=5)\n",
    "            ax.legend().remove()\n",
    "\n",
    "            ax.set_title(d_reg_num_name_icci[_rgi_reg], fontweight='bold', fontsize=24)\n",
    "\n",
    "            pd_proj_2100_2300 = pd_proj_2100_2300.dropna()\n",
    "            assert len(pd_proj_2100_2300) == 60\n",
    "            for j,sim_year in enumerate([2100,2300]):\n",
    "                # choose a \"frac\" parameter and select the right fitting values \n",
    "                frac, _sel = select_frac_w_experts(sel_it, sim_year=sim_year)\n",
    "                frac_dict[f'{rgi_reg}_{sim_year}'] = frac\n",
    "                \n",
    "                ax_fit = axs_fit[j]\n",
    "                ax_fit.set_title(f'{d_reg_num_name_icci[_rgi_reg]}:  year {sim_year}', fontweight='bold')\n",
    "\n",
    "                # get the raw data ...\n",
    "                x = pd_proj_2100_2300['global_temp_ch_2071-2100_preindustrial']\n",
    "                y = pd_proj_2100_2300[f'rel_ice_{sim_year}_{rgi_reg}']\n",
    "                \n",
    "                t_min, t_max = x.min(), x.max()\n",
    "                if sim_year == 2100:\n",
    "                    delta_x = 0.002\n",
    "                else:\n",
    "                    delta_x = 0.001\n",
    "\n",
    "                df_quantiles = _sel[['q50','q05','q25','q75','q95']]\n",
    "                # we just have to clip negative values .... there are still some although we tried to omit them by choosing the best frac ... \n",
    "                df_quantiles = df_quantiles.where(df_quantiles>=0, 0)\n",
    "                _sel_ice_eval = df_quantiles['q50'].values\n",
    "                _sel_ice_eval_bottom = df_quantiles['q05'].values\n",
    "                _sel_ice_eval_bottom_iqr = df_quantiles['q25'].values\n",
    "                _sel_ice_eval_top_iqr = df_quantiles['q75'].values\n",
    "                _sel_ice_eval_top = df_quantiles['q95'].values\n",
    "\n",
    "                ax_fit.plot(pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].deltaTemp,\n",
    "                         pd_fit_stats_reg_exp.loc[pd_fit_stats_reg_exp.year ==sim_year].q50, lw=3, color='green', alpha = 0.5, label='exponential fit')\n",
    "                ax_fit.plot(_sel.deltaTemp,\n",
    "                         _sel_ice_eval, lw=3, color='black', alpha = 1, label=f'lowess, frac={frac}')\n",
    "                ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom, _sel_ice_eval_top, lw=3, alpha=0.2, color=f'black', label='90% PI')\n",
    "                ax_fit.fill_between(_sel.deltaTemp, _sel_ice_eval_bottom_iqr, _sel_ice_eval_top_iqr, lw=3, alpha=0.4, color=f'black', label = '50% PI (IQR)')\n",
    "                \n",
    "                temp_eval_x = _sel.deltaTemp.values\n",
    "                sns.scatterplot(ax=ax_fit,\n",
    "                                data = pd_proj_2100_2300, #.loc[j], \n",
    "                                x='global_temp_ch_2071-2100_preindustrial', y=f'rel_ice_{sim_year}_{rgi_reg}',\n",
    "                                                style='model',\n",
    "                                hue='global_temp_ch_2071-2100_preindustrial',\n",
    "                                palette=cmap,      \n",
    "                                sizes=[150,150,150], size='model') #, legend=)\n",
    "                ax_fit.legend(fontsize=16, ncols=2)\n",
    "\n",
    "                ax_fit.set_xlabel('Temperature change between 1850-1900 & 2071-2100')\n",
    "                ax_fit.set_ylabel(f'% of Remaining Glacier Ice\\nin {sim_year} (relative to 2020)')\n",
    "\n",
    "                for j,temp_lev in enumerate(list(colors_temp_values.keys())):\n",
    "\n",
    "                    # for the annotation in the main plot ... \n",
    "                    if (temp_lev != '>4.0°C') and (temp_lev != '<=+4.0°C'):\n",
    "\n",
    "                        temp_lev_value = colors_temp_values[temp_lev]\n",
    "                        color= cmap(_scaler.transform(np.array([temp_lev_value]).reshape(-1,1))).flatten()\n",
    "                        idx_nearest = np.abs(temp_eval_x - temp_lev_value).argmin()\n",
    "                        if j >0:\n",
    "                            # check the previous fit, if they overlap, need to shift the text\n",
    "                            previous_fit = fit.copy()\n",
    "                        else:\n",
    "                            previous_fit = 10\n",
    "                            \n",
    "                            \n",
    "                        if previous_fit <4:\n",
    "                            shift = previous_fit-3.5\n",
    "                        else:\n",
    "                            shift = 0\n",
    "                        \n",
    "                        fit = _sel_ice_eval[idx_nearest]\n",
    "\n",
    "\n",
    "\n",
    "                        if sim_year == 2300:\n",
    "                            loc_2300_stats = 2300 +4 \n",
    "                        else:\n",
    "                            loc_2300_stats = 2100\n",
    "                        #if temp_lev_value>1.5:\n",
    "                        #    ax.vlines(loc_2300_stats,y_fitted[0], y_fitted[1], color=colors_temp[temp_lev], lw=15, alpha = 0.4)\n",
    "                        ax.hlines(fit,loc_2300_stats-3, loc_2300_stats+3, lw=8, color=color, zorder=10)\n",
    "\n",
    "                        if sim_year == 2300:\n",
    "                            bbox=dict(facecolor='white', edgecolor='white', boxstyle='round', alpha = 0.8)\n",
    "                            ax.annotate(temp_lev[2:], xy=(2300, fit),\n",
    "                                        xytext=(loc_2300_stats+delta_text, fit+shift),\n",
    "                                                    verticalalignment=\"center\",\n",
    "                                                    color=color, alpha=1, fontweight = 'bold', fontsize=18)\n",
    "                            _sel_temp_lev_qs = np.array([ _sel_ice_eval_bottom[idx_nearest],\n",
    "                                                         _sel_ice_eval_bottom_iqr[idx_nearest], _sel_ice_eval[idx_nearest],\n",
    "                                                         _sel_ice_eval_top_iqr[idx_nearest], _sel_ice_eval_top[idx_nearest]])\n",
    "                            if uncertainty == 'boxplot':\n",
    "                                boxplot_manual(qs= _sel_temp_lev_qs,\n",
    "                                               p = loc_2300_stats+7,\n",
    "                                               ax = ax, j=j*6, temp_val = temp_lev_value,\n",
    "                                               _scaler=_scaler,\n",
    "                                               )\n",
    "                np.testing.assert_allclose(temp_eval_x[::2], np.arange(1.4,6.3,0.1))\n",
    "                \n",
    "                lw = 3\n",
    "                if only_below45deg:\n",
    "                    temp_eval_x_sel = temp_eval_x[temp_eval_x<=4.5]\n",
    "                    _sel_ice_eval_sel = _sel_ice_eval[temp_eval_x<=4.5]\n",
    "                    colors_fit = cmap(_scaler.transform(temp_eval_x_sel[::2].reshape(-1,1))) # only every decile ... \n",
    "                    for c,f in zip(colors_fit, _sel_ice_eval_sel[::2]):\n",
    "                        ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                                   color=c, zorder=7)\n",
    "                else:\n",
    "                    colors_fit = cmap(_scaler.transform(temp_eval_x[::2].reshape(-1,1))) # only every decile ... \n",
    "                    for c,f in zip(colors_fit, _sel_ice_eval[::2]):\n",
    "                        ax.hlines(f,loc_2300_stats-1.5, loc_2300_stats+1.5, lw=lw, alpha = 0.5,\n",
    "                                   color=c, zorder=7)\n",
    "\n",
    "            ax.set_xlim([1997, xlim1])\n",
    "            if rgi_reg == 'RGI08':\n",
    "                ax_fit.set_ylim([-0.5, 120])\n",
    "\n",
    "            ax.set_xticks([2020,2100,2200,2300],[2020,2100,2200,2300])\n",
    "            ax.spines['right'].set_visible(False)\n",
    "            ax.spines['top'].set_visible(False)\n",
    "            \n",
    "            #ax_f.text(2290,119, caption_text,  #transform=ax_f.transAxes,\n",
    "            #            verticalalignment=\"top\", bbox=props, horizontalalignment='right',\n",
    "            #            color='grey', alpha=1, fontweight = 'bold', fontsize=18)\n",
    "            \n",
    "\n",
    "            fig.tight_layout()\n",
    "            if uncertainty== 'boxplot':\n",
    "                if add_median:\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/png/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_{vv}_final_{uncertainty}.png')\n",
    "                    fig.savefig(f'figures/main_figures_with_uncertainties/median_glac_models/svg/icci_report_2023_median_lowess_predi_oggm_glogem_pygem_temp_levels_{rgi_reg}_{vv}_final_{uncertainty}.svg')\n",
    "\n",
    "\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n",
    "            plt.close()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f81a7637",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "83671145",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "07a106e1",
   "metadata": {},
   "source": [
    "### Old some other checks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "0cb1e220",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.35 RGI16 -0.760643699402348 0.7600000000000005 -0.761960698358852 0.0020228495409952 0.0381786926738357\n",
      "0.95 RGI08 -0.0133381612731181 0.02 -0.0033727585556425 -0.0194944193027671 -0.1660472550030315\n",
      "0.8 RGI11 -0.2882249343312313 0.7900000000000005 -0.1665355601870256 -0.0166530028288747 -1.8135322844575616\n",
      "0.97 RGI02 -0.0668112543183777 0.20000000000000004 -0.0329958003550487 0.0942569718687187 -0.714569684132279\n",
      "0.41 RGI18 -1.1262397292353086 1.1200000000000008 -0.0165457309003222 -1.1262397292353086 0.0120141499051497\n",
      "0.8 RGI06 -0.1689932172761166 0.7500000000000004 -0.0615052678051349 0.924602637055692 -1.6090488460476096\n",
      "0.57 RGI13-14-15 -0.3423459518639633 0.4400000000000002 -0.3423459518639633 -0.0502908458109736 -0.0536211913880723\n",
      "0.54 RGI01 -0.0461755604138239 0.15 -0.0650186944139499 0.0266234330999974 0.4545621735883686\n",
      "0.97 RGI04 -0.0288227257040212 0.09999999999999999 -0.0012606868150747 0.2063598978217087 -0.4131474069450779\n",
      "0.43 RGI17 -0.3464151442898471 0.35000000000000014 -0.3568757993417772 -0.3225331531823912 0.0086394260179012\n",
      "0.37 global -0.18261132766947 0.18000000000000002 -0.1857867606668427 0.0056993210271052 0.454384797387198\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1400x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "                    \n",
    "sim_year = 2300\n",
    "for reg in list(d_reg_num_name_icci_reg.keys()):\n",
    "    rgi_regs = d_reg_num_name_icci_reg[reg]\n",
    "    #fig, axs = plt.subplots(3,2, figsize=(25,22)) #harey=True, sharex=True)\n",
    "    #fig_fit, axs_fit = plt.subplots(3,2, figsize=(25,22)) #harey=True, sharex=True)\n",
    "    #axs_fit = axs_fit.flatten()\n",
    "    #axs = axs.flatten()\n",
    "    len_sub_p = len(rgi_regs)\n",
    "\n",
    "    for jj,_rgi_reg in enumerate(rgi_regs): \n",
    "        list_index_l = []\n",
    "        err_l = []\n",
    "        for m in path_d.keys():\n",
    "            list_index_l.append(list(pd_cmip_temp_ch_2100.index + '_' + m))\n",
    "        pd_proj_2100_2300 = pd.DataFrame(index=np.concatenate(list_index_l))\n",
    "        #pd_proj_2100_2300 = pd.DataFrame(index=pd_cmip_temp_ch_2100.index)\n",
    "\n",
    "        if _rgi_reg !='global':\n",
    "            rgi_reg = f'RGI{_rgi_reg}'\n",
    "        else:\n",
    "            rgi_reg = _rgi_reg\n",
    "\n",
    "        pd_fit_stats_reg = pd_fit_stats.loc[pd_fit_stats.region == rgi_reg]\n",
    "        pd_fit_stats_reg_lowess = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'lowess_predi']\n",
    "        pd_fit_stats_reg_exp = pd_fit_stats_reg.loc[pd_fit_stats_reg.fit_opt == 'exp']\n",
    "        sel_it = pd_fit_stats_reg_lowess # we don't have it anymore ... .loc[pd_fit_stats_reg_lowess.it ==it]\n",
    "\n",
    "        diff_min = 0\n",
    "        repeat = True\n",
    "        while repeat:\n",
    "            _sel = sel_it.loc[sel_it.year == sim_year].copy()\n",
    "            _sel_a = sel_it.loc[sel_it.year == sim_year].copy()\n",
    "\n",
    "            if diff_min == 0:\n",
    "                plt.figure()\n",
    "                plt.plot(_sel.frac, _sel.min_q05_diff,'x')\n",
    "                plt.plot(_sel.frac, _sel.min_q25_diff,'x')\n",
    "                plt.plot(_sel.frac, _sel.min_q50_diff,'x')\n",
    "                plt.plot(_sel.frac, _sel.min_q75_diff,'x')\n",
    "                plt.plot(_sel.frac, _sel.min_q95_diff,'x')\n",
    "            \n",
    "            \n",
    "            # also make sure to select a non-negative median\n",
    "            # this does NOT work in some regions, we will just clip negative values afterwards ... \n",
    "            _sel = _sel.loc[(_sel['min_q50']>= -2) | (_sel['min_q50'] >= _sel['min_q50'].max())]\n",
    "            \n",
    "            # first choose only those that have the lowest fraction \n",
    "            min_q_diff = _sel[['min_q05_diff','min_q25_diff','min_q50_diff','min_q75_diff','min_q95_diff']].min(axis=1)\n",
    "            _sel.loc[min_q_diff.index,'min_q_diff'] = min_q_diff.values\n",
    "            max_min_q_diff = _sel['min_q_diff'].max()\n",
    "\n",
    "            _sel = _sel.loc[_sel['min_q_diff'] >= max_min_q_diff-0.1]\n",
    "            \n",
    "            #if rgi_reg == 'global':\n",
    "            #    print(_sel['min_q50'].unique())\n",
    "            if diff_min == 0:\n",
    "                plt.title(rgi_reg)\n",
    "                plt.plot(_sel.frac, _sel.min_q05_diff,'o')\n",
    "                plt.plot(_sel.frac, _sel.min_q25_diff,'o')\n",
    "                plt.plot(_sel.frac, _sel.min_q50_diff,'o')\n",
    "                plt.plot(_sel.frac, _sel.min_q75_diff,'o')\n",
    "                plt.plot(_sel.frac, _sel.min_q95_diff,'o')\n",
    "                plt.axhline(max_min_q_diff)\n",
    "            _sel = _sel.loc[_sel.min_q50_diff>=-0.01-diff_min/5]\n",
    "            _sel = _sel.loc[_sel.min_q95_diff>=-0.01-diff_min]\n",
    "            _sel = _sel.loc[_sel.min_q75_diff>=-0.01-diff_min]\n",
    "            _sel = _sel.loc[_sel.min_q25_diff>=-0.01-diff_min]\n",
    "            _sel = _sel.loc[_sel.min_q05_diff>=-0.01-diff_min]\n",
    "\n",
    "            _val = _sel.min_q50.max() if _sel.min_q50.max() <=0 else 0\n",
    "            _sel = _sel.loc[_sel.min_q50 >=_val]\n",
    "            _val = _sel.min_q05.max() if _sel.min_q05.max()<=0 else 0\n",
    "            _sel = _sel.loc[_sel.min_q05 >=_val]\n",
    "            if len(_sel) == 97:\n",
    "                frac = np.round(_sel.frac.unique(),2)[0]\n",
    "                repeat = False\n",
    "\n",
    "            elif len(_sel)>97:\n",
    "                frac = np.round(_sel.frac.unique().max(),2)\n",
    "                repeat = False\n",
    "            else:\n",
    "                repeat = True\n",
    "                diff_min = diff_min + 0.01\n",
    "                if diff_min >=5:\n",
    "                    break\n",
    "                #    proble\n",
    "                #    #frac = np.round(_sel.frac.unique(),2)[0]\n",
    "                #    print('problem', rgi_reg, sim_year, #frac,\n",
    "                #          diff_min)\n",
    "        plt.axvline(frac, lw=3, color='black')\n",
    "        print(frac, rgi_reg, max_min_q_diff, diff_min, _sel.min_q05_diff.max(), _sel.min_q95_diff.max(), _sel.min_q50.min())\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b3524575",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "oggm_v16",
   "language": "python",
   "name": "oggm_v16"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}